KL Workshop PDF

Part I
3G Overview
1
Company Confidential
What’s New in WCDMA?
Characteristic to WCDMA
• RAKE receiver takes advantage of multipath propagation
• Fast power control keeps system stable by using minimum power necessary for
links
• Soft handover ensures smooth handovers
Multiservice Environment Air Interface

• Data speed • Capacity and
– In RAN1 (Rel99)bit rate varies from 8 kbps up to coverage coupled -
384 kbps “cell breathing”
– Variable bit rate also available • Neighbor cells
– In RAS5.0 introduction of HSDPA; RAS06 HSUPA coupled via
interference
• Service delivery type
• Soft handover
– Real-time (RT) & non real-time (NRT)
• Fast power control
• Quality classes for user to choose
• Interference limited
– Different error rates and delays
system (e.g. GSM
• Traffic asymmetric in uplink & downlink frequency limited)
• Common channel data traffic (FACH)
• Inter-system handovers 2
UMTS network architecture Gi Gp
PSTN
GMSC GGSN
AuC
Gc
C H
HLR Gn
PSTN PSTN
Gr
D EIR
F Gf
G
VLR VLR
Gs
B B SGSN
MSC MSC
E
CN
A Gb IuCS IuPS
BSS RNS
Iur
BSC RNC RNC
Abis Iubis
BTS BTS Node B Node B

cell
Um Uu
ME
Microsoft Word SIM-ME i/f or Cu
Document
SIM USIM
Ref. 3GPP TS23.002 MS 3
3G Spectrum Allocation
4
IMT2000 Frequency Allocation for
UMTS
1900 1920 1980 2010 2025 2110 2170 2200
TDD FDD MSS TDD FDD MSS

UL/DL UL UL UL/DL DL DL MHz
FDL
FDL/UL
FUL
FDD Mode TDD Mode 5

• IMT 2000 3G Terms
– Third generation mobile systems as defined by ITU
– Global recommendation
• 3GPP
– 3rd Generation Partnership Project (Forum for a WCDMA standardization)
– Involved: ETSI (Europe), ARIB (Japan), TTA (Korea), T1P1 (USA), TTC (Japan) and
CWTS (China)
• 3GPP2
– 3rd Generation Partnership Project (Forum for a CDMA2000)
• UMTS
– Third generation telecommunication system, that is subject to specifications produced
by 3GPP
• WCDMA
– Air Interface technology adapted for UMTS Terrestrial Radio Access (UTRA)
• UTRA-FDD
– WCDMA in 3GPP, FDD mode
• UTRA-TDD
– WCDMA in 3GPP, TDD mode
• CDMA2000
– Air Interface technology proposal from TR45.5 (USA) on evolution of IS-95 (CDMA)
• TD-SCDMA
– Time Division Synchronous CDMA (TD-SCDMA) was proposed by China Wireless
Telecommunication Standards group (CWTS) and approved by the ITU in 1999
• MSS
6
– Mobile Satellite System
3G Standards
7
UMTS System Characteristics
• W-CDMA : 5 MHz
• Carrier Spacing : multiples of 200 kHz
• W-CDMA spreading rate = 3.84 Mchip/s
• Chip Rate = 3.84 MHz
• Raised cosine filtering with roll-off 0.22
• Information bit rate: between 8 kbit/s and 2 Mbit/s (currently up to 384 Kbit/s)
• Spreading Factor (SF): 4 -256
• Multiple Access Scheme : Wideband DS-CDMA
• Duplex Scheme : FDD and TDD modes
• Carrier Spacing : 4.4 – 5.4 MHz
• 10 ms frame with 15 time slots
• NodeB synchronization: asynchronous
• Highly variable data rates, data rate constant within 10 ms frame
• Bandwidth on demand, efficient resource usage
• Multiple services with different variable data rates over one physical channel
• DL Transmission diversity
8
Key features of WCDMA
•Soft handoff: user equipment (UE) and base stations use special rake receivers that
allow each UE to simultaneously communicate with multiple base stations. The diversity
gain associated with soft handoff is known as the "soft handoff gain factor".
– Soft handover (links from different NodeB) is performed at RNC level using Selection
Combination (RNC choose the best link)
– Softer handover (links from same NodeB) is performed at NodeB using Maximum Ratio
Combination (NodeB combines the signals). Softer handover usually is higher then soft
•Multi-path reception: the rake receivers also allow the UE to decode multiple signals
that have traveled over different physical paths from the base station. For example, one
signal may travel directly from the base station to the UE, and another may reflect off a
large building and then travel to the UE. This phenomenon, "multi-path propagation", also
provides a diversity gain. The same effect occurs on the uplink from the UE to the base
station.
•Power control: transmissions by the UE must be carefully controlled so that all

transmissions are received with roughly the same power at the base station. If power
control is not used, a “near-far” problem, where mobiles close to the base station over-
power signals from mobiles farther away, occurs. The base station uses a fast power
control system to direct the mobile to power up or power down as its received signal level
varies due to changes in the propagation environment. Likewise, on the downlink,
transmissions from the base stations are power-controlled to minimize the overall
interference throughout the system and to ensure a good received signal by the UE.
9
Key features of WCDMA
Frequency reuse of 1: every base station in the CDMA system operates on the same
frequency for a given carrier, so no frequency planning is required. As every site causes
interference to every other site, careful attention must be paid to each site's radio
propagation.
Soft capacity: capacity and coverage are intertwined in CDMA, depending on the
number of users in the system and the amount of interference allowed before access is
blocked for new users. By setting the allowed interference threshold lower, coverage will
improve at the expense of capacity. By setting the threshold higher, capacity will increase
at the expense of coverage. Because of the fundamental link between coverage and
capacity, cells with light traffic loads inherently share some of their latent capacity with
more highly loaded surrounding cells.
10
WCDMA Compared to GSM and CDMA IS-95
WCDMA vs. GSM
WCDMA has some similarities with GSM technology, however, it is a fundamentally
different technique for allowing multiple users to share the same spectrum and as a
result it has many differences.
11
WCDMA Compared to GSM and IS-95 CDMA
12
TD-SCDMA System Characteristics
Time Division Synchronous CDMA (TD-SCDMA) was proposed by China Wireless Telecommunication Standards
group (CWTS) and approved by the ITU in 1999 and technology is being developed by the Chinese Academy of
Telecommunications Technology and Siemens. TD-SCDMA uses the Time Division Duplex (TDD) mode, which
transmits uplink traffic (traffic from the mobile terminal to the base station) and downlink traffic (traffic from the
base station to the terminal) in the same frame in different time slots. That means that the uplink and downlink
spectrum is assigned flexibly, dependent on the type of information being transmitted. When asymmetrical data
like e-mail and internet are transmitted from the base station, more time slots are used for downlink than for
uplink. A symmetrical split in the uplink and downlink takes place with symmetrical services like telephony.
• TD-SCDMA Technical Summary

Frequency band: 2010 MHz - 2025 MHz in China (WLL 1900 MHz - 1920 MHz)
Minimum frequency band required: 1.6MHz
Frequency re-use: 1 (or 3)
Chip rate: 1.28 Mcps
Frame length: 10ms
Number of slots: 7
Modulation: QPSK or 8-PSK
Voice data rate: 8kbit/s
Circuit switched services: 12.2 kbits/s, 64 kbits/s, 144 kbits/s, 384 kbits/s, 2048 kbits/s
Packet data: 9.6kbits/s, 64kbits/s, 144kbits/s, 384kbits/s, 2048kbits/s
Receiver: Joint Detection, (mobile: Rake)
Power control period: 200 Hz
Number of slots / frame: 7
Frame length: 5ms
Multi carrier option
Handovers: Hard
Smart antennas
Uplink synchronization 13
Physical layer spreading factors: 1, 2, 4, 8, 16Company Confidential
QoS for different services
Real time
Non
Real time
14
RLC Transfer Mode
• AC determines the UL & DL RLC transfer mode for RNC & UE
• RLC transfer mode indicates, whether the RLC entity of a RB shall
use an Acknowledged (AM) (delivery guaranteed), an
Unacknowledged (UM) (delivery not guaranteed ) or a Transparent
(Tr) Mode (no protocol overhead added to higher layers ) data
transfer.
• Equal transfer mode is used in UL and DL for a RB.
• The transfer mode is determined from the RAB parameters and CN
domain information.
CS domain PS domain
Conversational Streaming Interactive Background
TM
TM TM
TM UM
UM AM
AM AM
AM AM
AM
RLC
RLC RLC
RLC RLC
RLC RLC
RLC RLC
RLC RLC
RLC
DCH
DCH DCH
DCH DCH
DCH DCH
DCH RACH/FACH
RACH/FACHDCH/HS-DSCH
DCH/HS-DSCH DCH/DCH
DCH/DCH
15
RT traffic NRT traffic
Conversational services
• Speech service:
– Real time conversational service require the low time delay from
end to end , and the uplink and the downlink service bandwidth
is symmetrical
– Adopt AMR ( adaptive multi rate ) technique (WCDMA).
• 12.2 (GSM), 10.2, 7.95, 7.40(IS-41), 6.70(PDC), 5.90, 5.15 and
4.75kbps
• The bit rate of AMR voice can be controlled by the RAN according
to the payload of air interface and the quality of voice service
• According to the requirement of the operator ,AMR technique can
balance the relationship among the network capacity, coverage and
the service quality
• Video phone (WCDMA)

– The requirement of time delay is similar to the voice service.
– The CS connection :adopt ITU-T Rec.H.324M
– The PS connection :adopt IETF SIP or H.323
16
Streaming services
• Multimedia data streaming:
– Preserve time relation between information entities of the data
streaming
– Data is processed into stable and continuous streaming
– Non-symmetry service
• Services example :
– Telemetry ( monitoring )
17
Interactive services
• Interactive traffic - fundamental characteristics for QoS:

– request response pattern
– preserve payload content
• Services example :
– Location based services
– Online game
– Web browsing
18
Background services
• Background traffic - fundamental characteristics for QoS:

– the destination is not expecting the data within a certain time
– preserve payload content
• Service example:
– E-mail ( server to server )
– SMS
– Download of database
– Reception of measurement records
19
Transmission diversity - STTD
• Space Time Transmit Diversity
20
Transmission diversity - TSTD
Time Switch Transmit Diversity, used in Synchronization physical channels
P-SCH and S-SCH
P-SCH
S-SCH
P-SCH
S-SCH
21
Part II
WCDMA Fundamentals
22
WCDMA = DS-CDMA
•WCDMA is a code-division multiple access technology which separates each
user’s voice or data information by multiplying the information by pseudo-random
bits called "chips".
•Direct sequence (DS): with DS, a binary modulated signal is ‘directly’ multiplied
by a code. The code is a pseudo-random sequence of ±1, where the bit rate of
the code (Chip Rate) is higher than the rate of the signal, usually considerably
higher. This has the effect of spreading the signal to a wideband. At the receiver,
the same code is used to extract the original signal from the incoming wideband
signal. A bit of the code is referred to as a chip, and the defining parameter for
such a system is the chip rate.
•The pseudo-random bit sequences have a rate of 3.84 Mcps (millions of chips
per second), resulting in the narrowband information bits of the user being spread
across a much wider bandwidth of approximately 5 MHz.
• For this reason, CDMA technology is sometimes referred to as “spread
spectrum.”
•Spread spectrum describes any system in which a signal is modulated so that its
energy is spread across a frequency range that is greater than that of the original
signal
•The user data (signal) is first spread by the channelisation code (based on
Hadamard matrix) called Orthogonal Variable Spreading Factor (OVSF) Code.
•OVSF code has the property that two different codes from the family are
perfectly orthogonal if in phase
23
TDMA based System
24
W-CDMA based System
25
Processing Gain and Spreading
According to information theory, as the frequency spectrum a signal occupies is expanded, the overall power level
decreases. In CDMA, the user signals are spread up to a wideband by multiplication by a code.
Consider a narrowband signal, say, for example, a voice call.
When viewed in the frequency spectrum, it occupies some frequency and has some power level. Once the frequency
is spread across a wideband, the total power of this signal is substantially reduced.
Now consider that another user has the same procedure performed on it and is also spread to the same wideband.
The total system power is increased by a small amount as the two users are transmitted at the same time. Therefore,
each new user entering the system will cause the power of the wideband to increase. 26
Spreading and Despreading
• At the receiver, the process of extracting one user is performed (DESPREADING)
• The regenerated signal needs to be retrieved with enough power that it can be
perceived above the level of the remaining spread signals. That is, it needs to be
of a sufficient strength, or margin, above the rest of the signals so that the signal
can be accurately interpreted.
• Considering this as a signal to interference ratio (SIR), or carrier to interference
(C/I) ratio, the noise affecting one signal is the remaining spread signals that are
transmitting at that frequency. This SIR is classified in CDMA as Eb/No.
• For mobile device measurements of the quality of the signals from the network, it
uses a pilot channel (CPICH), which is broadcast by each cell. The mobile device
measures Ec/Io, the energy level of this pilot channel, Ec, compared to the total
energy received, Io
• Another important characteristic is the rejection of unwanted narrowband noise
signals. If a wideband signal is affected by a narrowband noise signal, then since
the spreading function is commutative, the despreading operation while extracting
the wanted signal will in turn spread the narrowband noise to the wideband, and
reduce its power level
•The lower the power that the original signals are transmitted with, the lower the
noise in the system. It is therefore essential that each user in the system transmits
with an optimum power level to reach the receiver with its required power level. If
the power level is too high, then that user will generate noise, which in turn affects
the performance of all the other users. If there is too little power, then the signal
which reaches the receiver is of too low quality, and it cannot be accurately ‘heard’
27
• There are two solutions to the problem of noise levels:
– First, an Admission Control policy is required that monitors the number of users
and the noise level, and once it reaches some maximum tolerable level, refuses
admission of further users. In a cellular system, such admission control needs to
be considered not only for one cell, but also for the effects that noise levels within
that cell have on neighboring cells.
– The second solution is to implement Power Control. Each user needs to transmit
with just enough power to provide a clear signal at the receiver above the noise
floor. This should be maintained regardless of where the users are located with
respect to the receiver, and how fast they are moving. Power control needs to be
performed frequently to ensure that each user is transmitting at an optimum
level.
• The ratio of the original signal to the spread signal is referred to as the
spreading factor and is defined as:
Spreading factor (SF) = chip rate/symbol rate
WCDMA TDMA (GSM)

5 MHz, 1 carrier 5 MHz, 25 carriers
28
• In the next slides, the SF is 4. Hence, variable data rates can be supported by using
variable length codes and variable SF to spread the data to a common chip rate
• When considering CDMA systems, it is useful to define how the different signals
interact with each other. Correlation is defined as the relationship or similarity
between signals. For pulse-type waveforms, such as CDMA codes, the cross-
correlation between two signals is defined as:
where R12 is the correlation between two signals v1 and v2, and τ is their relative
time offset.
• For the code to be effective, the receiver must know the specific code (in this case
1,-1,1,-1, see next slides) which is being used for transmission and it must also be
synchronized with this transmission. On reception the receiver can then simply
reintroduce the correct code which is multiplied with the incoming signal and
reproduce the actual symbol sent by the transmitter.
• The receiver also needs to know the actual number of chips that represent a
symbol (spreading factor) so that the chips can be regenerated to the sent symbol
through averaging the value of the chips over the symbol time. This is achieved
through integration, where the chips are summed over the total time period of the
symbol they represent.
• The principle of correlation is used at the receiver to retrieve the original signal out
of the noise generated by all the other users’ wideband signal.
29
• At the receiver, the received signal is multiplied by the code and the result is
integrated over the period of each baseband bit to extract the original data.
Since the receiver has four chips over which to integrate, the procedure yields
a strong result at the output.
• However, consider now that the receiver does not know the correct code.
Then the integration process will result in a signal which averages to around
zero.
• For both of these, the relative strength of the desired signal and the rejection
of other signals is proportionate to the number of chips over which the
receiver has to integrate, which is the SF. Large SFs result in more
processing gain and hence the original signals do not need so much
transmission power to achieve a target quality level.
• As can be seen, the longer the symbol time (i.e. lower data rate and higher
chip rate), the longer the integration process, thus the higher the amplitude of
the summed signal. This is referred to as processing gain (Gp) and is directly
proportional to the SF used.
• For example, if the symbols were spread over 8 chips then the Gp will be 8; if
spread over 16 chips, Gp would be 16. This means that the processing gain is
higher for lower data rates than for higher data rates, i.e. lower data rates can
be sent with reduced power since it is easier to detect them at the receiver 30
SF= chip rate/symbol rate = 4
The spreading sequences must have good correlation properties to facilitate the
separation of the wanted signal from all others:
•One sharp and dominant peak of the autocorrelation function for zero phase shift
•As small as possible values of the autocorrelation function for all out-of-phase shift
•As small as possible values of the cross-correlation (different signals) function for
31
all phase shift Company Confidential
32
CDMA Multiple Access Advantages :
Multiple Access Features
1. All Users’ Signals overlap in TIME and FREQUENCY
2. Correlating the Received Signal despreads ONLY the WANTED SIGNAL
p p
S1
RECEIVER of USER 1
S1xC1
f f p
p S1 = S1 X C1 X C1
p p S2 X C2 X C1
S2
f
f
S2xC2
f f
33
CDMA Multiple Access Advantages :
Interference Rejection
p p
S1
S1xC1
f f
p
S1
p p
I I IxC1
f f
Correlation Narrowband Interference Spread the power

Only a small portion of the interfering signal energy passes the filter and
remain as residual interference 34
CDMA Principles
Tc : Chip Rate of the PN Code
m1(t)
M1(f) Tb : Information rate (voice/data)
1 -1 1
f
1/Tb
Tb 2Tb 3Tb
c1(t)
C1(f)
f
1/Tb 1/Tc
Tc 4Tc
m1(t).c1(t)
C1(f)* M1(f)
f
1/Tb 1/Tc
35
Processing gain (Gp)
• Gp = Wc/Wi
• Where
– Wc: chip rate
– Wi: user data rate
Wc f
Wi
• The more processing gain the system has, the more the power of uncorrelated
interfering signals is suppressed in the despreading process
• Thus, processing gain can be seen as an improvement factor in the SIR
(Signal to Interference Ratio) of the signal after despreading
• Example: Voice AMR 12.2 Kbps Gp = 10*log(3840000/12200)= 25 dB
• After despreading the signal power has to be typically few dB above the
interference and noise: Eb/No = 5dB; therefore the required wideband signal-
to-interference ratio is 5dB – Gp = -20 dB.
• In other words, the signal power can be 20 dB under the interference and the
WCDMA receiver can still detect the signal
• Wideband signal-to-interference ratio is also called carrier-to-interference ratio:
C/I
• Thanks to spreading and desporeading, C/I can be much lower in WCDMA
than GSM (C/I = 9-12 dB)
36
Processing gain (Gp)
Voice user (12,2 kbit/s) (User data rate) x (spreading ratio)=

R const.=W=3,84 Mcps
Unspread
narrowband signal
Power density
Processing Gain
G=W/R=25 dB
W
(W/Hz)
Spread wideband
signal
Frequency (Hz) •Spreading sequences of

different length
Packet data user (384 kbit/s) •Processing gain dependent
R
on user data rate
Processing Gain
Unspread G=W/R=10 dB
Power density
"narrowband"
W signal
(W/Hz)
Spread wideband
signal
Frequency (Hz)
37
Spreading in WCDMA
Consists of 2 operations:
1. Channelisation (OVSF: Orthogonal Variable SF)
• Transforms each symbol (data bit) to the number of chips (increases
bandwidth)
• Number of chips per symbol = Spreading Factor (SF)
2. Scrambling (does not affect the signal bandwidth)
• Scrambling code is applied (PN codes)
TX
Scrambling Code
Data Channel coding 3.84 Mcps 3.84 Mcps

(CRC, Encoder,
MOD
Interleaver,
Bit Rate convoluter, Rate Symbol Rate Chip Rate Chip Rate
Matching)
Channelisation code (OVSF)
38
Spreading in WCDMA
• Channel coding
– CRC attachment
• Check for error during transmission
• Voice: if CRC check returns error discard information
• Data: if CRC check returns error ask for retransmission
– Convolutional or Turbo Coding
• Convolution coding for voice and low speed signalling
• Turbo coding for large data transmission, better performances than
convolutional coding
– Interleaving
• Distribute error (burstly error) over data transmitted
– Rate Matching
• Match Symbol Rate to that accepted by spreading
• Rate matching techniques: Repeat or Puncturing (remove)
Example: voice 12.2 (SF 128)
Rs=3.84/128= 30Ksps
Rs 3.84 Mcps If the output of Channel cod is
Rb Channel
coding < 30Ksps, Rate matching
Symbol Rate Chip Rate repeats the info else uses
puncturing (remove) 39
(OVSF) Company Confidential
Spreading in WCDMA
12.2 Kbps Uplink Reference channel
64 Kbps Uplink Reference channel

40
OVSF properties
• In the spreading process, information symbols, which occupy a relatively narrow
bandwidth, are multiplied by a high-rate spreading code consisting of chips
• The resulting spread signal has a wider bandwidth dependent on the number of
chips per symbol
• In the de-spreading process, the spreading code is multiplied by the spread signal
to recover the original data symbols. The de-spreading process converts the wide
bandwidth spread signal back to the original narrower bandwidth of the data
symbols
• Spreading codes (OVSF) are specially designed to allow the symbols from
multiple users to occupy the same spectrum at the same time, while still allowing
the original information to be retrieved.
• Codes are allocated in RNC
• OVSF code has the property that two different
codes from the family are perfectly orthogonal
if in phase
• Restrictions: another physical channel may use a
certain code in the tree if no other physical channel
to be transmitted using the same code tree is using
a code that is on an underlying branch, i.e. using a
higher SF generated from the intended spreading
code to be used. Neither can a smaller SF code on
the path to the root of the tree be used SF4 41
OVSF properties
• The signals that are all being transmitted at the same time and frequency must be
separated out into those from individual users. This is the second role of the code.
Considering the party analogy, if this was a GSM party, then the problem is solved
easily. All guests must be quiet and each is then allowed to speak for a certain time
period; no two guests speak at the same time. At a CDMA party, all users are allowed to
speak simultaneously, and they are separated by speaking in different languages, which
are the CDMA codes. All of the codes that are used must be unique and have ideally no
relationship to each other. Mathematically speaking, this property is referred to as
orthogonality. The system can support as many simultaneous users as it has unique or
orthogonal codes.
• Orthogonal codes are used in CDMA systems to provide signal separation. As
long as the codes are perfectly synchronized, two users can be perfectly
separated from each other.
• To generate a tree of orthogonal codes, a Walsh–Hadamard matrix is used. For perfect
orthogonality between two codes, for example, it is said that they have a cross-
correlation of zero when τ = 0. Consider a simple example using the following two codes:
42
OVSF properties
• To verify if two codes have a zero cross-correlation, they are tested in the below equation, first
multiplied together and then integrated, as shown in Figure below. The result is zero, indicating that
indeed they are orthogonal.
• The number of chips which represent a symbol is known as the SF or the processing gain. To support
different data rates within the system, codes are taken from an appropriate point in the tree. These
types of orthogonal codes are known as orthogonal variable spreading factors (OVSF).
• In the 3G WCDMA system the chip rate is constant at 3.84 Mchips/s. However, the number of chips
that represent a symbol can vary. Within this system as laid down by the specifications, the minimum
number of chips per symbol is 4 which would give a data rate of 3 840 000/4 = 960 000 symbols per
second. The maximum SF or number of chips per symbol is 256,1 which would give a data rate of 3
840 000/256 = 15 000 symbols per second. Thus it can be seen that the fewer chips used to represent
a symbol, the higher the user data rate. The actual user data rate must be rate matched to align with
one of these SF symbol rates.
• Although orthogonal codes demonstrate perfect signal separation, they MUST be perfectly
synchronized to achieve this. Another drawback of orthogonal codes is that they do not evenly
spread signals across the wide frequency band, but rather concentrate the signal at certain
discrete frequencies. As an example, consider that the code ‘1 1 1 1’ will have no spreading
effect on a symbol. To overcome these drawbacks the PN codes are introduced
43
Scrambling code properties
• Another code type used in CDMA systems is the pseudo-random noise (PN)
sequence. This is a binary sequence of ±1 that exhibits characteristics of a
purely random sequence, but is deterministic. Like a random sequence, a PN
sequence has an equal number of +1s and −1s, with only ever a difference of
1. PN sequences are extremely useful as they fulfill two key roles in data
transmission:
1. Even spreading of data: when multiplied by a PN sequence, the resultant signal is
spread evenly across the wideband. To other users who do not know the code, this
appears as white noise.
2. Signal separation: while PN sequences do not display perfect orthogonality
properties, nevertheless they can be used to separate signals. At the receiver, the
desired signal will show strong correlation, with the other user signals exhibiting
weak correlation
• Another property of PN sequences is that they exhibit what is known as
autocorrelation. This is defined as the level of correlation between a signal and
a time-shifted version of the same signal, measured for a given time shift. For
a PN sequence, the autocorrelation is at a maximum value, N, when perfectly
time aligned, i.e. τ = 0. N is the length in numbers of bits of the PN sequence.
This single peak drops off quickly at ±Tc, where Tc is the width of a chip of the
code. This allows a receiver to focus in on where the signal is, without a
requirement for the transmitter and receiver to be synchronized. In
comparison, the autocorrelation of time-shifted orthogonal codes results in
several peaks, which means that this signal locking is much more problematic.44
Scrambling code properties
• The OVSF codes are effective only when the channels are perfectly
synchronized at symbol level
• The loss in cross-correlation, e.g. due to multipaths, is compensated
by the additional scrambling operation
• Scrambling codes are used to separate different cells in the downlink
and different terminals in the uplink
• They have good correlation properties (interference averaging) and
are always used on top of the spreading codes, thus not affecting the
transmission bandwidth
• Gold sequence is used to generate scrambling codes
• For downlink physical channels, a total of 218 = 262,143 scrambling
codes can be generated
• Only scrambling codes k = 0, 1, ..., 8191 are used
• 8192 scrambling codes are divided into 512 groups each of which
contains 16 scrambling codes
• The first scrambling code of each group is called Primary Scrambling
Code (PSC) and the other 15 are Secondary Scrambling Codes
(SSC)
45
Usage of the codes
Channelization Code Scrambling Code
Usage Uplink: separation of physical data Uplink: Separation of terminals
(DPDCH) and control channels (DPCCH)
for the same terminal
Downlink: separation of downlink Downlink: Separation of sectors
connections to different users within on cell (cells)
Length 4-256 chips Uplink: 10ms = 38400 chips
In downlink also 512 chips
Downlink: 10ms = 38400 chips
Number of codes Spreading Factor indicates the number of Uplink: over 16 millions
codes under one scrambling code Downlink: 512
Code Family Orthogonal Variable Spreading Factor 10ms code: Gold Code
(OVSF) 66.7µs code: Extended code
family
Spreading Yes, indicates bandwidth No, does not affect bandwidth
46
Receivers
• Both NodeB and Terminals use the same type of correlation
receivers
• Due to multipath propagation it’s necessary to use multiple
correlation receivers (fingers) in order to recover (combine) the
energy from all paths coherently and obtain multipath diversity
47
Multipath propagation and diversity
• A transmission from a mobile device is more or less omni directional, and this is also the
case for base stations when they have only one cell. Base stations which are sectorized will
have directional antennas, which will transmit only over a certain range.
• For example, a three-sectored site will have three antennas which each transmit over the
range of 120 degrees. From the point of view of the mobile device, it would be ideal if a
transmission were unidirectional; however, this is impractical since it would mean that the
antenna of the mobile device would need to point towards the base station at all times.
• In this ideal situation the device could transmit with reduced power, thus causing less
interference to other users and increasing the device’s battery life. In the cellular
environment, much of the power transmitted is actually in the wrong direction. In urban
areas there is considerable reflection of the signal from surrounding buildings. This is
actually a reason why cellular systems work, since the mobile device can thus be out of
direct line of sight of the BTS and its signal will still be received.
• The reflected signals travel further distances than the direct line of sight transmission and
therefore arrive slightly later,
later with greater attenuation and possible phase difference.
• It would be advantageous if these time-shifted versions in the multipath signal could be
combined at the receiver with the effect that a much stronger signal is received.
• Because this combined signal is stronger, it is possible that the BTS may tell the mobile
device to reduce its transmitting power. Any process of combining multiple versions of the
same signal to provide a more powerful, better quality signal is known as diversity.
• The autocorrelation property of the PN sequence is again used. Since the received signal
resolves into a single peak around the chip width, then as long as the multipath profile is of
a duration longer than the chip width, a number of peaks will be observed, each one
representing a particular multipath
48
Wide Band Channel
• Definition:
• A channel is defined wide when its bandwidth (Bw) is greater than the
Coherence Bandwidth: Bw >> ∆fc
1
∆f c =
2πSτ
49
Wide Band Channel – Delay Spread
Channel impulse response (power delay profile) and delay spread
Dominant Path
τ1
50
Wide Band Channel – Narrow/Wide Band System
Microsoft Word
Document
51
WCDMA and GSM in TU3 Channel
52
Optimal Receiver for WCDMA signal
• For a channel with only one signal path optimal receiver is one correlator
(code de-spreading and integration
Basic unit of
Rake Receiver
53
Optimal Receiver for WCDMA signal
• In a multipath environment optimal receiver utilizes several correlators (Rake
Fingers) tuned for dominant delays = Rake receiver Adobe Acrobat
Document
54
Rake Receiver
• Rake finger delays tuned based on channel impulse
response estimation
• Code Matched Filter, Search Finger
• Fingers combined with Maximal Ratio combining
• Performance of Rake Receiver depends on the channel
powers delay profile
• Max path delay difference vs. chip time amount of
multipath diversity
55
Rake Receiver - Combining
• Combined signal without and with phase estimation and correction (example
6 path channel)
56
Maximal Ratio Combining of Symbols
Transmitted Received signal Time and phase Combined

signal (+noise) adjustment signal (+
residual noise)
Finger n.1
Finger n.2
Finger n.3
UE
WBTS
57
Maximal Ratio Combining of Symbols
Transmitted Received Modified with channel Combined
symbol symbol+noise estimate and relative symbol +
delay compensation residual
(for combining) noise
Finger n.1
Finger n.2
Finger n.3
UE
WBTS
58
WCDMA in TU Channel
time
Corr 1 Corr 3
Corr 2 Corr 4 Corr 5 Corr 6 Corr 7
• High level of multipath diversity
59
WCDMA in Indoor Channel
Rake Finger RESOLUTION = 0.26µs 78m

Corr 1 Chip period = 1/3840000 s = 0.26µs
• No multipath diversity
60
Part III
Scrambling Code Planning
61
62
63
64
65
66
67
68
69
70
71
72
Part IV
Physical Layer
73
Channel Mapping
In GSM, we distinguish between logical and physical channels. In UMTS there are three different
types of channels:
• Logical Channels
Logical Channels were created to transmit a specific content. There are for instance logical channel to
transmit the cell system information, paging information, or user data. Logical channels are offered as data
transfer service by the Medium Access Control (MAC) layer to the next higher layer. Consequently, logical
channels are in use between the mobile phone and the RNC.
• Transport Channels (TrCH)
The MAC layer is using the transport service of the lower, the Physical layer. The MAC layer is responsible to
organise the logical channel data on transport channels. This process is called mapping. In this context, the
MAC layer is also responsible to determine the used transport format. The transport of logical channel data
takes place between the UE and the RNC.
• Physical Channels (PhyCH)
The physical layer offers the transport of data to the higher layer. The characteristics of the physical
transport have to be described. When we transmit information between the RNC and the UE, the physical
medium is changing. Between the RNC and the Node B, where we talk about the interface Iub, the
transport of information is physically organised in so-called Frames.
Between the Node B and the UE, where we find the WCDMA radio interface Uu, the physical transmission
is described by physical channels. A physical channel is defined by the UARFCN and the a spreading
code in the FDD mode.
74
Radio Interface Channel Organisation
Logical Channels (define what type of data is transferred ) RLC

content is organised in separate channels, e.g. Layer
System information, paging, user data, link management
L2
Transport Channels
define how and with which type of characteristics the MAC
data is transferred by the physical layer Layer
Physical Channels Frames PHY

L1 (UARFCN, spreading code) Iub interface Layer
75
Radio Interface Protocol Architecture GC Nt DC
Duplication avoidance
Nt DC L3 RRC (Radio Resource Control): Broadcast of information

GC
provided by the non-access stratum (Core Network). Broadcast
UuS boundary
of information related to the access stratum Establishment, re-
C-plane signalling U-plane information establishment, maintenance and release of an RRC connection
between the UE and UTRAN , Establishment, reconfiguration
and release of Radio Bearers, Assignment, reconfiguration and
release of radio resources for the RRC connection, RRC
RRC control L3 connection mobility functions, Paging/notification, Control of
requested QoS, UE measurement reporting and control of the
reporting, Outer loop power control
Radio
control
control
control
control
Bearers L2/PDCP (Packet Data Convergence Protocol): Header

compression and decompression, Support for lossless SRNS
PDCP relocation
PDCP L2/PDCP
L2/BMC (Broadcast/Multicast Control): Storage of Cell broadcast

BMC
L2/BMC messages, dealing with CBS (Cell Broadcast Servises)
L2/RLC (Radio Link Control): Segmentation (Reassembly),

RLC
Retransmission, Error correction [Transparent (no overhead
RLC L2/RLC
RLC added, e.g. voice/video), Unack (add overhead but no
RLC
RLC RLC retransmission, e.g. cell broadcast), Ack (add overhead and
RLC RLC retransmission enable, e.g. PS)]
Logical L2/MAC (Medium Access Control): Mapping Logical channels

Channels to transport Reporting of measurements.Chipering. Local
measurements such as traffic volume and quality indication are
MAC L2/MAC reported to RRC
Transport L1:Closed loop PC, Macrodiversity distribution/combining and soft
Channels handover execution; Error detection on transport channels and indication
to higher layers; FEC encoding/decoding and interleaving/deinterleaving
PHY L1 of transport channels; Multiplexing of transport channels and
demultiplexing of coded composite transport channels; Rate matching;
Mapping of coded composite transport channels on physical
channels; Power weighting and combining of physical channels;
Modulation and spreading/demodulation and despreading of
physical channels; 76
Frequency and time (chip, bit, slot, frame) synchronisation;
Company Confidential Measurements and indication to higher layers (e.g. FER, SIR,
interference power, transmit power, etc.);
Logical Channels
There are two types of logical channels (FDD mode):
Control Channels (CCH):
• Broadcast Control Channel (BCCH)
System information is made available on this channel. The system information informs the UE about the
serving PLMN, the serving cell, neighbourhood lists, measurement parameters, etc.
This information permanently broadcasted in the downlink.
• Paging Control Channel (PCCH)
Given the BCCH information the UE can determine, at what times it may be paged. Paging is required,
when the RNC has no dedicated connection to the UE. PCCH is a downlink channel.
• Common Control Channel (CCCH)
Control information is transmitted on this channel. It is in use, when no RRC connection exists between
the UE and the network. It is a bi-directional channel, i.e. it exists both uplink and downlink.
• Dedicated Control Channel (DCCH)
Dedicated resources were allocated to a UE. These resources require radio link management, and the
control information is transmitted both uplink and downlink on DCCHs.
Traffic Channels (TCH):
• Dedicated Traffic Channel (DTCH)
User data has to be transferred between the UE and the network. Therefore dedicated resources can be
allocated to the UE for the uplink and downlink user data transmission.
• Common Traffic Channel (CTCH)
Dedicated user data can be transmitted point-to-multipoint to a group of UEs.
77
Transport Channels (TrCH)
Logical Channels are mapped onto Transport Channels. There are two types of Transport
Channels (FDD mode):
Common Transport Channels:

• Broadcast Channel (BCH)
It carries the BCCH information.
• Paging Channel (PCH)
It is in use to page a UE in the cell, thus it carries the PCCH information. It is also used to
notify UEs about cell system information changes.
• Forward Access Channel (FACH)
The FACH is a downlink channel. Control information (FACH-c), but also small amounts of
user data can be transmitted on this channel (FACH-u).
• Downlink Shared Channel (DSCH)
This channel is used downlink. Dedicated user data and control information for several mobile
phones can be transmitted with one DSCH.
• Random Access Channel (RACH)
This uplink channel is used by the UE, when it wants to transmit small amounts of data, and
when the UE has no RRC connection. It is often used to allocated dedicated signalling
resources to the UE to establish a connection or to perform higher layer signalling. It is a
contention based channel, i.e. several UE may attempt to access UTRAN simultaneously.
78
Transport Channels (TrCH)
• Common Packet Channel (CPCH)
Similar to the RACH, it is a contention based uplink channel. In contrast to the RACH, it can
be used to transmit larger amounts of (bursty) traffic.
Dedicated Transport Channels:

• Dedicated Channel (DCH)
Dedicated resources can be allocated both uplink and downlink to a UE. Dedicated resources
are exclusively in use for the subscriber.
On the following figures. you can see the mapping of logical channels onto transport channels,
as well as the mapping of transport channels onto physical channels.
79
Physical Channels (PhyCH)
Physical Channels are characterised by
•UARFCN,
•scrambling code,
•channelisation code (optional),
•start and stop time, and
•relative phase (in the uplink only, with relative phase being 0 or π/2)
Transport channels can be mapped to physical channels. But there exist physical channels, which
are generated at the Node B only, as can be seen on the next figures.
In addition to the physical channels mapped from the transport channels, there exist
physical channels for signaling purposes (blue color) to carry only information
between network and the terminals.
80
Channel Mapping DL (Network Point of View)
Logical Transport Physical
Channels Channels Channels
P-CPICH
BCCH BCH P-CCPCH
PCCH PCH S-CCPCH

PICH
CCCH AICH
FACH
P-SCH
CTCH
S-SCH
DCCH HS- HS-PDSCH
DSCH HS-SCCH
DTCH DCH DPDCH
DPCCH
81
Channel Mapping UL (Network Point of View)
Logical Transport Physical
CCCH RACH PRACH
DCCH
HS-DPCCH
DTCH DCH DPDCH I branch
DPCCH Q branch
82
Example – Channel configuration during call
Data Logical Transport Physical
RRC
DCCH0-4 DCH1
signalling
DPDCH
Speech
DTCH1 DCH2-4 DPCCH
data
AMR speech
NRT
data
DTCH2 DCH5 +
NRT data
AMR speech connection utilises multiple transport channels
RRC connection utilises multiple logical channels, signalling radio bearers
83
Transport Channels Physical Channels
• Transport channels contain the data generated at the higher layers,

which is carried over the air and are mapped in the physical layer to
different physical channels.
• The data is sent by transport block (TB) from MAC layer to physical
layer and generated by MAC layer every 10 ms (TTI)
• The transport format of each transport channel is identified by the
Transport Format Indicator (TFI), which is used in the interlayer
communication between the MAC layer and physical layer.
• Several transport channels can be multiplexed together by physical
layer to form a single Coded Composite Transport Channel
(CCTrCh).
84
Transport Formats
TFCS
TB
DCH 2
TB TB TB
TTI TTI TTI
TB
TB
TBS DCH 1
TB TB
TB TB TB
TFS
TTI TTI TTI
TFC
TF
TB Transport Block TF Transport Format

TBS Transport Block Set TFS Transport Format Set
TFC Transport Format Combination
TFCS Transport Format Combination Set 85
Cell Synchronisation
When a UE is switched on, it starts to monitor the radio interface to find a suitable cell to camp on.
But it has to determine, whether there is a WCDMA cell nearby. If a WCDMA cell is available, the
UE has to be synchronised to the downlink transmission of the system information – transmitted on
the physical channel P-CCPCH – before it can make a decision, in how far the available cell is
suitable to camp on. Initial cell selection is not the only reason, why a UE wants to perform cell
synchronisation. This process is also required for cell re-selection and the handover procedure.
Cell synchronisation is achieved with the Synchronisation Channel (SCH). This channel divides up
into two sub-channels, P-SCH and S-SCH are not under the cell-specific primary scrambling
code (the UE must be able to synchronize to the cell before knowing the downlimk scrambling
code)
•Primary Synchronisation Channel (P-SCH) (SLOT and CHIP SYNCHRONIZATION)

A time slot lasts 2560 chips. The P-SCH only uses the first 10% of a time slot. A Primary
Synchronisation Code (PSC) is transmitted the first 256 chips of a time slot. This is the case in every
UMTS cell. If the UE detects the PSC, it has performed TS and chip synchronisation.
•This is typically done with a single matched filter matched to the primary synchronization code
which is common for all cells. The slot timing of the cell can be obtained by decoding peaks in
the matched filter output
Matched 86
(continued on the next text slide) Filter
Synchronisation Channel (SCH)
2560 Chips 256 Chips

Primary Synchronisation Channel (P-SCH)
CP CPP CP CP
Secondary Synchronisation Channel (S-SCH)
Cs1 Cs2 Cs15 Cs1
Slot 0 Slot 1 Slot 14 Slot 0
10 ms Frame
Cp = Primary Synchronisation Code (Activity Factor 10%)
Cs = Secondary Synchronisation Code (Activity Factor 10%)
87
Cell Synchronisation
• Secondary Synchronisation Channel (S-SCH)
• (FRAME SYNCH and Scrambling Code Group DETECTION)
The S-SCH also uses only the first 10% of a timeslot; Secondary Synchronisation Codes (SSC) are
transmitted. There are 16 different SSCs, which are organised in a 10 ms frame (15 timeslots) in
such a way, that
• the beginning of a 10 ms frame can be determined, and
• 64 different SSC combinations within a 10 ms frame are identified. There is a total of
512 primary scrambling codes, which are grouped in 64 scrambling code families, each
family holding 8 scrambling code members. The 15 SSCs in one 10 ms frame identify the
scrambling code family of the cell‘s downlink scrambling code.
• The sequence permits downlink frame synchronization and indicate which of the code
grouping the downlink scrambling code belongs to.
• This is done by correlating the received signal with all possible secondary
synchronization code sequences and identifying the maximum correlation value. Since
the cyclic shifts of the sequences are unique, the code group as well as the frame
synchronization is determined
88
SSC Allocation for S-SCH
scrambling slot number
code group 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
group 00 1 1 2 8 9 10 15
15 8 10 16 2 7 15 7 16
group 01 1 1 5 16 7 3 14 16 3 10 5 12 14 12 10
group 02 1 2 1 15 5 5 12 16 6 11 2 16 11 15 12
group 03 1 2 3 1 8 6 5 2 5 8 4 4 6 3 7
group 04 1 2 16 6 6 11 15 5 12 1 15 12 16 11 2
group 05 1 3 4 7 4 1 5 5 3 6 2 8 7 6 8
group 62 9 11 12 15
15 12 9 13 13 11 14 10 16 15 14 16
group 63 9 12 10 15 13 14 9 14 15 11
11 11 13 12 16 10
I monitor
the S-SCH
11 15 5
89
Common Pilot Channel (CPICH)
With the help of the SCH, the UE was capable to perform chip, TS, and frame synchronisation. Even
the cell‘s scrambling code group is known to the UE. But in the initial cell selection process, it does
not yet know the cell‘s primary scrambling code. There is one primary scrambling code in use over the
entire cell, and in neighbouring cells, different scrambling codes are in use. There exists a total of 512
primary scrambling codes.
The CPICH is used to transmit in every TS a pre-defined bit sequence (stream of 256 ‘1’) with a fixed
data rate of 30 kbps, which corresponds to spreading factor 256. The CPICH divides up into a
mandatory Primary Common Pilot Channel (P-CPICH) and optional Secondary CPICHs (S-CPICH).
The P-CPICH is in use over the entire cell. And it is the first physical channel, where a spreading code
is in use. A spreading code is the product of the cell‘s scrambling code and the channelisation code.
The channelisation code is fixed: Cch,256,0. I.e., the UE knows the P-CPICH‘s channelisation code, and it
uses the P-CPICH to determine the cell‘s primary scrambling code by trial and error (UE tries 8 SC
Codes of the group identified).
The P-CPICH is not only used to determine the primary scrambling code. It also acts as
- phase reference for most of the physical channels,
- measurement reference in the FDD mode (and partially in the TDD mode).
There may be zero or several S-CPICHs. Either the cell‘s primary scrambling code or its secondary
scrambling codes can be used. In contrast to the P-CPICH, it can be broadcasted just over a part of the
cell.
CPICH has activity factor of 100% (continuous transmission)
90
Primary Common Pilot Channel (P-CPICH)
10 ms Frame
CP
P-CPICH
Cell scrambling
code? I get it with
trial & error! P-CPICH
applied spreading code =
cell‘s primary scrambling code ⊗ Cch,256,0
• Phase reference
• Measurement reference
91
CPICH as Measurement Reference
The UE has to perform a set of L1 measurements, some of them refer to the CPICH channel:
• CPICH RSCP
RSCP stands for Received Signal Code Power. The UE measures the RSCP on the Primary-
CPICH. The reference point for the measurement is the antenna connector of the UE. The
CPICH RSCP is a power measurement of the CPICH. The received code power may be high, but
it does not yet indicate the quality of the received signal, which depends on the overall noise
level.
• UTRA carrier RSSI.
RSSI stands for Received Signal Strength Indicator. The UE measures the received wide band
power, which includes thermal noise and receiver generated noise. The reference point for the
measurements is the antenna connector of the UE.
• CPICH Ec/No
The CPICH Ec/No is used to determine the „quality“ of the received signal. It gives the received
energy per received chip divided by the band‘s power density. The „quality“ is the primary
CPICH‘s signal strength in relation to the cell noise. (Please note, that transport channel
quality is determined by BLER, BER, etc. )
If the UE supports GSM, then it must be capable to make measurements in the GSM bands,
too. The measurements are based on the
• GSM carrier RSSI
The wideband measurements are conducted on GSM BCCH carriers.
92
P-CPICH as Measurement Reference
CPICH RSCP Received Signal Code Power (in dBm)
CPICH Ec/No received energy per chip divided by the power density in the band (in dB)
UTRA carrier received wide band power, including thermal noise and noise
RSSI generated in the receiver (in dBm)
CPICH Ec/No = CPICH RSCP

UTRA carrier RSSI
CPICH Ec/No CPICH RSCP UTRA carrier RSSI

0: -24 0: -115 0: -110
1: -23.5 1: -114 1: -109
2: -23 2: -113 2: -108
3: -22.5 : :
... 88: -27 71: -39
47: -0.5 89: -26 72: -38
48: 0 73: -37
Ec/No values in dB RSCP values in dBm RSSI values in dBm

93
Primary Common Control Physical Channel (P-CCPCH)
The UE knows the cell‘s primary scrambling code. It now wants to gain the cell system
information (MIB,SIB), which is transmitted on the physical channel P-CCPCH. The
channelisation code of the P-CCPCH is also known to the UE, because it must be Cch,256,1 in every
cell for every operator. By reading the cell system information on the P-CCPCH, the UE learns
everything about the configuration of the remaining common physical channels in the cell, such
as the physical channels for paging and random access.
As can be seen from the P-CCPCH‘s channelisation code, the data rate for cell system
information is fixed. The SCH is transmitted on the first 256 chips of a timeslot, thus creating
here a peak load. The cell system information is transmitted in the timeslot except for the first
256 chips. By doing so, a high interference and load at the beginning of the timeslot is avoided.
This leads to a net data rate of 27 kbps for the cell system information.
Channel estimation is done with the CPICH, so that no pilot sequence is required in the P-
CCPCH. (The use of the pilot sequence is explained in the context of the DPDCH later on in this
document.) There are also no power control (TPC) bits transmitted to the UE‘s.
P-CCPCH has activity factor of 90%
94
10 ms Frame
CP
P-CCPCH
Finally, I get the

cell system
information P-CCPCH
• channelisation code: Cch,256,1
• no TPC, no pilot sequence
• 27 kbps (due to off period)
• organised in MIBs and SIBs
95
In GSM all common channels have the same power,

there is no need to consider the power setting of
common channels as all common channels are on full
power. In GSM we have to decide how many TSLs to
dedicate to common tasks, I.e. how many SDCCH TSL
are required per cell
96
Nokia Parameters for Cell Search
• WCEL: PtxPrimaryCPICH
The parameter determines the transmission power of the primary CPICH
channel. It is used as a reference for all common channels.
[-20 dBm … 43 dBm], step 1 dB, default: 33dBm (WPA power = 43 dBm)
• WCEL: PtxPrimarySCH
Transmission power of the primary synchronization channel, the value is
relative to primary CPICH transmission power.
[-35 dB … 15 dB], step size 0.1 dB, default: -3 dB
• WCEL: PtxSecSCH
Transmission power of the secondary synchronization channel, the value is
[-35 dB… 15 dB], step size 0.1 dB, default: -3 dB
97
Nokia Parameters for Cell Search
• WCEL: PtxPrimaryCCPCH
This is the transmission power of the primary CCPCH channel, the value is
[-35 dB … 15 dB], step size 0.1 dB, default: -5 dB
• WCEL: PriScrCode
Identifies the downlink scrambling code of the Primary CPICH (Common Pilot
Channel) of the Cell.
[0 ... 511], default: 0 dB
98
Secondary Common Control Physical Channel (S-CCPCH)
The S-CCPCH can be used to transmit the transport channels
• Forward Access Channel (FACH) and

• Paging Channel (PCH).
More than one S-CCPCH can be deployed. The FACH and PCH information can multiplexed on
one S-CCPCH – even on the same 10 ms frame -, or they can be carried on different S-CCPCH’s.
When 2 S-CCPCH’s are broadcast, the first S-CCPCH has a spreading factor of 256 and
carries PCH, while the spreading factor of the remaining S-CCPCH can range between 256 (30
Kbps or 15 Ksps) and 4 (1920 Kbps) and carries FACH. UTRAN determines, whether a S-CCPCH
has the TFCI (Transport Format Combination Indicator) included (supports variable rates). Please
note, that the UE must support both S-CCPCHs with and without TFCI.
S-CCPCH is on air ONLY when there is data to transmit (FACH or Paging), however the TFCI bits
are broadcast irrespective of whether or not there is any data to transmit (min activity factor
25%)
Typical value is SF = 64 120 Kbps (60 Ksps)
99
Secondary CCPCH Power Setting
• The Secondary CCPCH (Common Control Physical Channel) carries

FACH and PCH transport channels
• Parameter NbrOfSCCPCHs (Number of SCCPCHs) tells how many
SCCPCHs will be configured for the cell. (1, 2 or 3)
• If only one SCCPCH is used in a cell, it will carry FACH-c (Containing
DCCH/CCCH /BCCH), FACH-u (containing DTCH) and PCH. FACH and
PCH multiplexed onto the same SCCPCH.
• If two SCCPCHs are used in a cell, the first SCCPCH will always carry PCH
only and the second SCCPCH will carry FACH-u and FACH-c.
• If three SCCPCHs are used in a cell, the third SCCPCH will carry FACH-s
(containing CTCH) and FACH-c idle (containing CCCH and BCCH ) . The
third SCCPCH is only needed when Service Area Broadcast (SAB) is active
in a cell.
100
• DL common Channel configuration in case of three SCCPCH For SAB
Logical channel DTCH DCCH CCCH BCCH CTCH PCCH
Transport FACH-u FACH-c FACH-c FACH-s PCH

channel
Physical channel SCCPCH SCCPCH SCCPCH

connected idle page
SF 64 SF 128 SF 256
101
• The power of SCCPCHs are set relative to CPICH transmission power, but it is
based on the bitrate.
– The Spreading factor for SCCPCH, which is carrying FACH (with or without
PCH), is 64 (60ksps)
– The Spreading factor for SCCPCH, which is carrying PCH only is 256 (15ksps)
– The Spreading factor for SCCPCH, which is carrying FACH-s/FACH-c idle for
SAB, is 128 (30ksps)
• Recommended value of the transmission power of the SCCPCH is depended on the

number of SCCPCHs . Default transmission power values for different SCCPCHs
are as follow:
– PtxSCCPCH1 (SF=64) (default 0) For PCH/FACH or standalone PCH
– PtxSCCPCH2 (SF=256) (default –5) For Standalone PCH
– PtxSCCPCH3 (SF=128) (default –2) For SAB
102
Secondary Common Control Physical Channel (S-CCPCH)
10 ms Frame
TFCI
Data Pilot bits
(optional)
• carries PCH and FACH

• Multiplexing of PCH and FACH on one
S-CCPCH, even one frame possible
• with and without TFCI (UTRAN set) S-CCPCH
• SF = 4..256
• (18 different slot formats)
• no inner loop power control
• a maximum of 1 paging message
can be sent per 10 ms TTI
103
S-CCPCH and the Paging Process
• The network has detected, that there is data to be transmitted to the UE (MTC). Both in the
RRC idle mode and in the RRC connected mode (e.g. in the sub-state CELL_PCH) a UE may get
paged. But how does the mobile know, when it was paged ? And in order to save battery
power, we don‘t want the UE to listen permanently to paging channel – instead, we want to
have discontinuous reception (DRX) of paging messages. But WHEN and WHERE does the UE
listen to the paging messages?
• Cell system information is broadcasted via the P-CCPCH. The cell system information is
organised in System Information Blocks (SIB). SIB5 informs the mobile phones about the
common channel configuration, including a list of S-CCPCH descriptions. The first 1 to K
entries transmit the (transport channel) PCH, while the remaining S-CCPCH in the list hold no
paging information.
• The UE determines the S-CCPCH, where it is paged, by its IMSI and the number of PCH
carrying S-CCPCHs K. When paging the UE, the RNC knows the UE‘s IMSI, too, so that it can
put the paging message on the correct PCH transport channel.
• Discontinuous Reception (DRX) of paging messages is supported. A DRX cycle length k has to
be set in the network planning process for the cs domain, ps domain, and UTRAN.
k ranges between 3 and 9. If for instance k=6, then the UE can be paged every 2k = 640
ms. If the UE is in the idle mode, it takes the smaller k-value of either the cs- or ps-domain. If
the UE is in the connected mode, it has to select the smallest k-value of UTRAN and the CN, it
is not connected to.
104
S-CCPCH and the Paging Process
UTRAN
BCCH (SIB 5)
common
channel
definition,
UE including Node B RNC
a lists of
Index of S-CCPCHs
0 S-CCPCH carrying one PCH
1 S-CCPCH carrying one PCH
K-1 S-CCPCH carrying one PCH
S-CCPCH without PCH

UE‘s paging channel:
Index = IMSI mod K S-CCPCH without PCH
„my paging
e.g. if IMSI mod K = 1 channel“ 105
The Paging Process
Paging Indicator Channel (PICH)
UMTS provides the terminals with an efficient sleep mode operation. The UEs do not have to
read and process the content, transmitted during their paging occasion on their S-CCPCH.
Each S-CCPCH, which is used for paging, has an associated Paging Indicator Channel (PICH). A
PICH is a physical channel, which carries paging indicators. A set of (paging indicator) bits
within the PICH indicate to a UE, whether there is a paging occasion for it. Only then, the UE
listens to the S-CCPCH frame, which is transmitted 7680 chips after the PICH frame in order
to see, whether there is indeed a paging message for it.
The PICH is used with spreading factor 256. 300 bits are transmitted in a 10 ms frame, and
288 of them are used for paging indication (activity factor 96%). The UE was informed by the
BCCH, how many paging indicators exist on a 10 ms frame. The number of paging indicator
Np can be 18, 32, 72, and 144, and is set by the operator as part of the network planning
process. The higher Np, the more paging indicators exist, the more paging groups exist, among
which UEs can be distributed on. Consequently, the lower the probability, that a UE reacts on
a paging indicator, while there is no paging message in the associated S-CCPCH frame
(saving battery consumption). But a high number of paging indicators results in a
comparatively high output power for the PICH (increase DL interference), because less bits
exists within a paging indicator to indicate the paging event. The operator then also has to
consider, if he has to increase the number of paging attempts.
106
S-CCPCH and its associated PICH
S-CCPCH frame,
τS-CCPCH
associated with PICH frame
τPICH
= 7680
PICH frame chips
10 ms
for paging indication no transmission
b0 b1 b286 b287 b288 b299
# of paging Subscribers with Subscribers with

indicators per frame Pq indicator Pq indicator
(Np) paged => not paged =>
18 (16 bits) {b16q, …,b16q+15} = {1,1,…,1} {b16q, …,b16q+15} = {0,0,…,0}
32 (8 bits) {b8q, …, b8q+7} = {1,1,…,1} {b8q, …, b8q+7} = {0,0,…,0}
72 (4 bits) {b4q, …, b4q+3} = {1,1,…,1} {b4q, …, b4q+3} = {0,0,…,0} 107
144 (2 bits) {b2q, b2q+1} = {1,1}
Company Confidential {b2q, b2q+1} = {0,0}
Nokia Parameters for S-CCPCH and Paging
RAN 1 & RAN1.5 support data rates of 15, 30, and 60 ksym/s for the S-CCPCH. FACH Open
Loop power control can be implemented only if the S-CCPCH is dedicated, uplink PC
information through the RACH (RAN 2)
• WCEL: NbrOfSCCPCHs
The parameter defines how many S-CCPCH are configured for the given cell.
Range: [1,2], step: 1; default = 1 (1 = FACH&PCH; 2 = FACH on 1st / PCH on 2nd)
• WCEL: PtxSCCPCH1 (carries FACH & PCH)
This is the transmission power of the 1st S-CCPCH channel, the value is relative
to primary CPICH transmission power.
Range: [-35 dB … 15 dB] , step size 0.1 dB, default: - 5dB
• WCEL: PtxSCCPCH2 (carries PCH only)
This is the transmission power of the 2nd S-CCPCH channel, the value is
Range: [-35 dB … 15 dB] , step size 0.1 dB, default: - 5dB
108
Nokia Parameters for S-CCPCH and Paging
• WCEL: PtxPICH
This is the transmission power of the PICH channel. It carries the paging
indicators
which tell the UE to read the paging message from the associated secondary
CCPCH. This parameter is part of SIB 5.
[-10 dB..5 dB]; step 1 dB; default: -8 dB (with Np =72)
NP
Repetition of PICH bits
[18, 36, 72, 144] with relative power [-10, -10, -8, -5] dB
• RNC: CNDRXLength
The DRX cycle length used for CN domain to count paging occasions for
discontinuous reception. This parameter is given for CS domain and PS domain
separately. This parameter is part of SIB 1.
[640, 1280, 2560, 5120] ms; default = 640 ms.
• WCEL: UTRAN_DRX_length
The DRX cycle length used by UTRAN to count paging occasions for
discontinuous reception.
[80, 160, 320, 640, 1280, 2560, 5120] ms; default = 320 ms
109
FACH and S-CCPCH
• The transport channel Forward Access Channel (FACH) is used, when relatively small amounts of data have to
be transmitted from the network to the UE. In-band signalling is used to indicate, which UE is the recipient
of the transmitted data (see MAC PDU with UE-ID type). This common downlink channel is used without
(fast) closed loop power control and is available all over the cell. FACH data is transmitted in one or several
S-CCPCHs. FACH and PCH data can be multiplexed on one S-CCPCH, but they can also be be transmitted on
different S-CCPCHs.
• The FACH is only transmitted downlink. The FACH is organised in FACH Data Frames via the Iub-interface.
Each FACH Data Frames holds the Transmission Blocks for one TFS. The used TFS is identified by the TFI. A TFI
is associated with one Transmission Time Interval (TTI), which can be either 10, 20, 40 or 80 ms. The TTI
identifies the interleaving time on the radio interface.
A FACH Data Frame has header fields, which identify the CFN, TFI, and the Transmit Power Level.
• The Transmit Power Level gives the preferred transmission power level for the FACH and for the TTI time. The
values specified here range between 0 and 25.5 dB, with a step size of 0.1 dB. The value is taken as a
negative offset to the maximum power configured for the S-CCPCHs, specified for the FACH.
• The pilot bits and the TFCI-field may have a relative power offset to the power of the data field, which may
vary in time. (The offset is determined by the network.) The power offsets are set by the NBAP message
COMMON TRANSPORT CHANNEL SETUP REQUEST, which is sent from the RNC to the Node B. There are two
power offset information included:
• PO1:defines the power offset for the TFCI bits; it ranges between 0 and 6dB with a 0.25 step size.
• PO3:defines the power offset for the pilot bits; it ranges between 0 and 6dB with a 0.25 step size.
Another important parameter is the maximum allowed power on the FACH: MAX FACH Power.
110
FACH and S-CCPCH
Power Transmit Power Level

offsets for FACH Data Frame
TFCI and TPC CFN TFI TB TB
defined
during Iub
channel Uu
setup
Node B RNC
max. transmit
power for S-CCPCH
UE
0..25.5 dB,
step size 0.1
Transmit Power Level PO1 PO3

Pilot bits
TFCI
Data
(optional)
111
Nokia Parameters for S-CCPCH Power Setting
Currently, either one or two S-CCPCHs are supported.
• WCEL: PowerOffsetSCCPCHTFCI
Defines the power offset for the TFCI symbols relative to the downlink
transmission power of a Secondary CCPCH. This parameter is part of SIB 5.
P01_15/30/60
15 ksps: [0..6 dB]; step 0.25 dB; default: 2 dB
30 ksps : [0..6 dB]; step 0.25 dB; default: 3 dB
• WCEL: PowerOffsetSCCPCHPilot
Defines the power offset for the pilot symbols relative to the downlink
transmission power of a Secondary CCPCH. This parameter is part of SIB 5.
P03_15/30/60
112
Code Tree Capacity
Note: there are not P-SCH and S-SCH !! P-SCH and S- S-SCH
are not under the cell-
cell- specific primary scrambling code
(the UE must be able to synchronize to the cell before
knowing the downlimk scrambling code)
There are 5 CCH's (4 use

SF256 and one uses SF64),
they blocked (5 used + 13
not allowed) a total of 18
codes
113
Part V
Power Control
114
Effect of TX & RX Powers on Interference Levels
Downlink transmission power

= Interference to the network
Uplink transmission power Uplink received power =

= Interference to other cells Interference to own cell users
Since every TX and RX power is causing interference to others, PC is

necessary to limit the interference 115
CDMA Fundamentals : Power Control
Near-Far Problem
Pr,1 = EIRP(MS1) - PL1 = 21 - 100 = -79 dBm

Pr,2 = EIRP(MS2) - PL2 = 21 - 90 = -69 dBm PL2 = 90 dB
Pr,1
Pr,2
P = 21 dBm
P = 21 dBm
PL1 = 100 dB
MS2
MS1
(S/N)1 = Pr,1 - Pr,2 = -10 dB MS2 must be Power Controlled by -10 dB to have
(S/N)2 = Pr,2 - Pr,1 = +10 dB the same S/N for both users MS1 and MS2
116
Near-Far Effect
117
Purpose of Power Control in WCDMA
118
Physical Random Access (Open loop Power Control)
Outer Loop Power Control
Fast Closed Loop (Inner) Power Control
119
Physical Random Access (Open loop Power Control)
In the random access (based on Slotted ALOHA approach with fast acquisition indication) ,
initiated by the UE (MOC), two physical channels are involved:
• Physical Random Access Channel (PRACH)

The physical random access is decomposed into the transmission of preambles in the uplink.
Each preamble is transmitted with a higher output power as the preceding one. After the
transmission of a preamble, the UE waits for a response by the Node B. This response is sent
with the physical channel Acquisition Indication Channel (AICH), telling the UE, that the Node
B has acquired the preamble transmission of the random access. Thereafter, the UE sends the
message itself, which is the RACH/CCCH of the higher layers. The preambles are used to allow
the UE to start the access with a very low output power. If it had started with a too high
transmission output power, it would have caused interference to the ongoing transmissions in
the serving and neighbouring cells.
Please note, that the PRACH is not only used to establish a signalling connection to UTRAN. It
can be also used to transmit very small amounts of user data.
• Acquisition Indication Channel (AICH)

This physical channel indicates to the UE, that it has received the PRACH preamble and is now
waiting for the PRACH message part.
120
Random Access – the Working Principle
UE Node B
No response PRACH (preamble)
by the
Node B
No response PRACH (preamble)

by the
Node B
I just detected
PRACH (preamble) a PRACH preamble
OLA!
AICH
PRACH (message part)
121
Random Access Timing
The properties of the PRACH are broadcasted (SIB5, SIB6). The candidate PRACH is randomly
selected (if there are several PRACH advertised in the cell) as well as the access slots (= 2 TIME
SLOTS) within the PRACH. 15 access slots are given in a PRACH, each access slot lasting two
timeslots or 5120 chips. In other words, the access slots stretch over two 10 ms frames.
frames A
PRACH preamble, which is transmitted in an access slot, has a length of 4096 chips.
Also the AICH is organised in (AICH) access slots, which stretch over two timeslots. AICH access
slots are time aligned with the P-CCPCH. (Activity factor 80%)
The UE sends one preamble in uplink access slot n. It expects to receive a response from the
Node B in the downlink (AICH) access slot n, τp-a chips later on. If there is no response, the UE
sends the next preamble τp-p chips after the first one. The maximum numbers of preambles in
one preamble access attempt can be set between 1 and 64. The number of PRACH preamble
cycles can be set between 1 and 32.
If the AICH_Transmission_Timing parameter in the SIB is set to BCCH SIB5 & SIB6 to
•0,
then, the minimum preamble-to-preamble distance is 3 access slots, the minimum preamble-to-message
distance is 3 access slots, and the preamble-to-acquisition indication is 3 timeslots.
•1,
then, the minimum preamble-to-preamble distance is 4 access slots, the minimum preamble-to-message
distance is 4 access slots, and the preamble-to-acquisition indication is 5 timeslots.
122
Random Access Timing
SFN mod 2 = 0 SFN mod 2 = 1 SFN mod 2 = 0
P-CCPCH
AICH access
slots 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 0 1 2 3 4 5 6 7
5120
chips
(distances depend on AICH_Transmission_Timing )
UE point of view
Acquisition
AICH Indication
access slots
preamble-to-AI AS # i
distance τp-a
4096 chips
PRACH Preamble Message

access slots Preamble part
5120 chips AS # i
preamble-to-preamble preamble-to-message
123
distance τp-p distance τp-m
PRACH Power Setting
Preamble_Initial_Power =
UL interference
1st preamble: + Primary CPICH TX power – CPICH_RSCP
power setting + Constant Value
Constant Value
attenuation
estimated receive level
in the DL
UL interference
at Node B -5..10
dB
“RRC
Connection
1..8 dB Request”
Pp-p Pp-m
Pp-p
Pre- Message
Pre- part
Pre- amble
amble
amble
# of preambles: 1..64 # of preamble cycles: 1..32 124

Nokia Parameters Related to the PRACH and AICH
WCEL: PRACHRequiredReceivedCI
This UL required received C/I value is used by the UE to calculate the initial output
power on PRACH according to the Open loop power control procedure. This
parameter is part of SIB 5.
[-35 dB..-10 dB]; step 1 dB; default -25 dB. We use - 20
WCEL: PowerRampSteponPRACHPreamble
UE increases the preamble transmission power when no acquisition indicator is
received by UE in AICH channel. This parameter is part of SIB 5.
[1dB..8dB]; step 1 dB; default: 2 dB. We use 1
• WCEL: PowerOffsetLastPreamblePrachMessage
The power offset between the last transmitted preamble and the control part of the
PRACH message.
[-5 dB..10 dB]; step 1 dB; default 2dB
• WCEL: PRACH_preamble_retrans
The maximum number of preambles allowed in one preamble ramping cycle, which is
part of SIB5/6.
[1 ... 64]; step 1; default 8. We use 7
125
• WCEL: RACH_tx_Max
Maximum number of RACH preamble cycles defines how many times the PRACH
pre-amble ramping procedure can be repeated before UE MAC reports a failure on
RACH transmission to higher layers. This message is part of SIB5/6.
[1 ... 32]; default 8. We use 16
WCEL: PRACHScramblingCode
The scrambling code for the preamble part and the message part of a PRACH
Channel, which is part of SIB5/6.
[0 ... 15]; default 0.
• WCEL: AllowedPreambleSignatures
The preamble part in a PRACH channel carries one of 16 different orthogonal
complex signatures. Nokia Node B restrictions: A maximum of four signatures can be
allowed (16 bit field).
[0 ... 61440]; default 15. We use 4
• WCEL: AllowedRACHSubChannels
A RACH sub-channel defines a sub-set of the total set of access slots (12 bit field).
[0 ... 4095]; default 4095.
126
• WCEL: PtxAICH
This is the transmission power of one Acquisition Indicator (AI) compared to CPICH
power. This parameter is part of SIB 5.
[-22 ... 5] dB, step 1 dB; default: -8 dB.
• WCEL: AICHTraTime
AICH transmission timing defines the delay between the reception of a PRACH
access slot including a correctly detected preamble and the transmission of the
Acquisition Indicator in the AICH.
0 ( Delay is 0 AS), 1 ( Delay is 1 AS) ;default 0.
• WCEL: RACH_Tx_NB01min
In case that a negative acknowledgement has been received by UE on AICH a backoff
timer TBO1 is started to determine when the next RACH transmission attempt will be
started. The backoff timer TBO1 is set to an integer number NBO1 of 10 ms time
intervals, randomly drawn within an Interval 0 ≤ NB01min ≤ NBO1 ≤ NB01max (with
uniform distribution).
[0 ... 50]; default: 0.
• WCEL: RACH_Tx_NB01max
[0 ... 50]; default: 50.
127
“RRC Conn
Setup”
UL DPDCH “RRC
Conn Setup
Complete”
“RRC Conn
Request”
OL PC is needed to keep the quality of the communication at the required level (BLER,
SIR, BER,…) by setting the target (SIR) for the fast power control. It aims at providing the
required quality: no worse, no better. Too high quality would waste capacity. It is needed
in both UL and DL since there is Fast PC (Closed Loop or Inner Loop) in both UL and DL
128
In RADIO BEARER SETUP Message you can find the Target BLER (for the DL)
For AMR and PS 128 = 1% BLER, CS T (VIDEO) = 0.1%, CS NT = 0.2%
129
UL Outer Loop Power Control Algorithm
Case of Soft Handover
130
UL Outer Loop Power Control Algorithm
When Max SIR Target is hit,

RNC might force a hard
handover
131
UL OL PC: BLER Eb/No (Initial SIR Target, SIR Target Max, SIR Target Min)
132
DL Outer Loop Power Control
DeltaSIR(1,2), DeltaSIR after (1,2),…..
The adjustments of the SIR Target done by the UE is a proprietary algorithm that
provides the same measured quality (BLER) as the quality target set by the RNC
133
134
• UL (Near-Far Problem): UE1 and UE2 operate within the same frequency,
separable at the base station only by their respective spreading codes. It
may happen that UE1 at the cell edge suffers a path loss, say 70 dB above
that of UE2 which is near to NodeB. If there were no mechanism for UE1
and UE2 to be power-controlled to the SAME level at the NodeB, UE2 could
easily overshoot UE1 and thus a large part of the cell. Power control tries
to equalizes the Rx power per bit of all UE’s at NodeB. Since Fast Fading is
uncorrelated between uplink and downlink (large freq separation between ul
and dl bands in FDD) we can not use only a method based on Open Loop
Power Control. Solution: Closed Loop PC: in UL the NodeB performs
frequent (1.5 KHz) estimates of the received SIR and compares it to the
SIR Target (calculated during Outer Loop PC).
• DL: We do not have Near-Far Problem due to one-to-many scenario: all the
signals within one cell originate from one NodeB to all mobiles. However it is
desirable to provide a marginal amount of additional power to UE at the cell
edge, as they suffer from increased other-cell-interference.
135
DL Fast Closed (Inner) Loop Power Control
Inner loop power control is also often called (fast) closed loop power control. It takes place between
the UE and the Node B. We talk about UL inner loop power control, when the Node B returns
immediately after the reception of a UE‘s signal a power control command to the UE. By doing so,
the UE‘s SIR ratio is kept at a certain level.
DL inner loop power control control is more complex. When the UE receives the transmission of the
Node B, the UE returns immediately a transmission power control command to the Node B, telling
the Node B either to increase or decrease its output power for the UE‘s DPCH. The Node B‘s
transmission power can be changed by 0.5, 1, 1.5 or 2 dB. 1 dB must be supported by the
equipment. If other step sizes are supported or selected, depends on manufacturer or operator.
The transmission output power for a DPCH has to be balanced for the PICH, which adds to the
power step size.
There are two downlink inner loop power control modes:
• DPC_MODE = 0: Each timeslot, a unique TPC command is sent uplink.
• DPC_MODE = 1: 3 consecutive timeslots (for DL), the same TPC command is transmitted.
One reason for the UE to request a higher output power is given, when the QoS target has not been
met. It requests the Node B to transmit with a higher output power, hoping to increase the quality
of the connection due to an increased SIR at the UE‘s receiver. But this also increases the
interference level for other phones in the cell and neighbouring cells. The operator can decide,
whether to set the parameter Limited Power Increase Used. If used, the operator can limit the
output power raise within a time period.
136
DL Fast Closed (Inner) Loop PC Algorithm
Every 1500 Hz (time slot)

UE measures SIR= (RSCP/ISCP)×SF
137
Downlink Inner Loop Power Control
TPC
two modes
cell
DPC_MODE = 0 DPC_MODE = 1
unique TPC command same TPC over 3 TS,

per TS then new command
1500 times/s 500 times/s TPCest per

1 TS / 3 TS
138
UL Inner Loop Power Control
SIRest
SIRtarget
time
T TC TCP
TC CP P = = 0
P= = 0 TPC ⇒
1 1
TPC_cmd
in FDD mode:
1500 times per second
139
UL Fast Closed (Inner) Loop PC Algorithm
140
Power Control Algorithm 1 is applied in medium speed environments. Here, the UE is
commanded to modify its transmit power every timeslot. If the received TPC value is 1, the UE
increases the transmission output at the DPCCH by ∆DPCCH, otherwise it decreases it by ∆DPCCH.
The ∆DPCCH is either 1 or 2 dB, as set by the higher layer protocols. TPC values from the same
radio link set represent one TLC_cmd. TPC_cmds from different radio link sets have to be
weighted, if there is no reliable interpretation.
Power Control Algorithm 2 (300 times/s) was specified to allow smaller step sizes in the
power control in comparison to PCA1. This is necessary in very low and high speed
environments. In these environments, PCA1 may result in oscillating around the target SIR.
PCA2 changes only with every 5th timeslot, i.e. the TPC_cmd is set to 0 (do not do anything)
the first 4 timeslots. In timeslot 5, the TPC_cmd is –1, 0, or 1. For each radio set (Radio set is
combined radio links from same NodeB), the TPC_cmd is temporarily determined. This can be
seen in the next figure. The temporary transmission power commands (TPC_temp) are
combined as can be seen in the figure after the next one. Here you can see, how the final
TPC_cmd is determined.
141
UL Inner Loop Power Control Algorithms (1 and 2)
• The optimum PC step size varies depending on the UE speed. For a given
quality target, the best UL PC step size is the one that results in the lowest
target SIR. With an update rate of 1500 Hz, a PC step size of 1dB can
effectively track a typical Rayleigh fading channel up to Doppler frequency of
about 55 Hz (30 Km/h). At higher speeds, up to about 80 Km/h, a PC step
size of 2dB gives better results.
• For speeds greater than 80 Km/h the inner loop PC can no longer follow the
fades and just introduces noise into the UL transmission. This adverse effect
on the UL performance could be reduced if a PC step size smaller than 1 dB
was employed. Also, for UE speeds lower than about 3 Km/h where the
fading rate of the channel is very small, a smaller step size is more
beneficial.
• Algorithm 1 is used when the UE speed is sufficiently low to compensate for
the fading of the channel (PC step size should be 1 or 2 dB)
• Algorithm 2 was designed for emulating the effect of using a PC step size
smaller than 1 dB and can be used to compensate for the slow fading trend
of the propagation channel rather than rapid fluctuations. It performs better
than Alg 1 when the UE moves faster than 80 Km/h or slower than 3 Km/h.
The UE does not change its transmission power until it has received 5
consecutive TPC commands. 142
algorithms for processing power
control commands TPC_cmd
PCA1 PCA2
TPC_cmd for each TS TPC_cmd for 5th TS
TPC_cmd values: +1, -1 TPC_cmd values: +1, 0, -1
step size ∆ TPC: 1dB or 2dB step size ∆ TPC: 1dB
UL DPCCH power adjustment: ∆DPCCH = ∆ TPC × TPC_cmd
PCA2 PCA1 PCA2
0 ≈3 ≈ 80 km/h
Rayleigh fading can be compensated
143
Soft Handover Case: UL Inner Loop Power Control Algorithm 1
Example: reliable transmission At the mobile, a “power down”

command has higher priority
Cell 3 over “power up” command
“UP” TPC3 = 1
⇒ “Down”
TPC_cmd = -1
TPC1 = 1 TPC3 = 0
“UP” “Down”
Cell 1
Cell 2
144
No Soft Handover Case - UL Inner Loop Power Control Algorithm 2 (Part 1)
TPC = 1 TPC_temp
TPC = 1 0
TPC = 1 0
0 • if all TPC-values = 1
TPC = 1
⇒ TPC_temp = +1
TPC = 1 0
• if all TPC-values = 0
TPC = 1 1
⇒ TPC_temp = -1
TPC = 0 0 • otherwise
TPC = 1 0 ⇒ TPC_temp = 0
TPC = 0 0
TPC = 1 0
TPC = 0 0
TPC = 0 0
TPC = 0 0
TPC = 0 0
TPC = 0 0
-1
145
Soft Handover Case :UL Inner Loop Power Control Algorithm 2 (Part 2)
Example:
N = 3 cells TPC_temp1 TPC_temp2 TPC_temp3
1 N
∑
N i=1
TPC_tempi
-1 -0.5 0 0.5 1
TPC_cmd = - 0 1 146
1 Company Confidential
Part VI
Dedicated Physical
Channels
147
Downlink Dedicated Physical Channel (DPCH)
The downlink DPCH is used to transmit the DCH data. Control information and user data are time
multiplexed. The control data is associated with the Dedicated Physical Control Channel (DPCCH),
while the user data is associated with the Dedicated Physical Data Channel (DPDCH). Data is not
only User Traffic/Data but could be also High layer signalling (“Measurement control”, “RRC
messages”,L3 Dedicated signalling)
The transmission is organised in 10 ms radio frames, which are divided into 15 timeslots. The
timeslot length is 2560 chips. Within each timeslot, following fields can be found:
• Data field 1 and data field 2, which carry DPDCH information
• Transmission Power Control (TPC) bit field
• Transport Format Combination Indicator (TFCI) field, which is optional
• Pilot bits
The exact length of the fields depends on the slot format, which is determined by higher layers. The
TFCI is optional, because it is not required for services with fixed data rates. Slot format are also
defined for the compressed mode; hereby different slot formats are in used, when compression is
archived by a changed spreading factor or a changed puncturing scheme. The pilot sequence is
used for channel estimation as well as for the SIR ratio determination within the inner loop power
control. The number of the pilot bits can be 2, 4, 8 and 16 – it is adjusted with the spreading
factor. A similar adjustment is done for the TPC value; its bit numbers range between 2, 4 and 8.
The spreading factor for a DPCH can range between 4 and 512. The spreading factor can be
changed every TTI period.
Superframes last 720 ms and were introduced for GSM-UMTS handover support.
148
Superframe = 720 ms
Radio Frame Radio Frame Radio Frame Radio Frame

0 1 2 71
10 ms Frame
TFCI
TPC
Data 1 bits bits Data 2 bits Pilot bits
bits
(optional)
DPDCH DPCCH DPDCH DPCCH

• 17 different slot formats
2,4,8,16 bits (SIR estimation,
• Compressed mode slot
format for changed SF &
phase estimation for Rake
2,4,8 bits receiver) 149
changed puncturing
Following features are supported in the downlink:
• Blind rate detection, and
• Discontinuous transmission.
Rate matching is done to the maximum bit rate of the connection. Lower bit rates are possible,
including the option of discontinuous transmission. Please note, that audible interference
imposes no problem in the downlink, since Common Channels have continuous transmission.
Multicode usage:
Several physical channels can be allocated in the downlink to one UE. This can occur, when
several DPCH are combined in one CCTrCH in the PHY layer, and the data rate of the CCTrCH
exceeds the maximum data rates allowed for the physical channels. Then, on all downlink
DPCHs, the same spreading factor is used. Also the downlink transmission of the DPCHs takes
place synchronous. One DPCH carries DPDCH and DPCCH information, while on the remaining
DPCHs, no DPCCH information is transmitted.
But also in the case, when several DPCHs with different spreading factors are in use, the first
DPCH carries the DPCCH information, while in the remaining DPCHs, this information is
omitted (discontinuous transmission).
Multicode usage is not implemented in RAN1.

150
Physical Layer Bit Rates (Downlink)
Spreading Channel Channel DPDCH Maximum user

factor symbol bit rate channel bit data rate with ½-
rate (kbps) rate range rate coding
(ksps) (kbps) (approx.)
512 7.5 15 3–6 1–3 kbps
256 15 30 12–24 6–12 kbps Half rate speech
128 30 60 42–51 20–24 kbps Full rate speech
64 60 120 90 45 kbps
32 120 240 210 105 kbps
16 240 480 432 215 kbps 128 kbps
8 480 960 912 456 kbps 384 kbps
4 960 1920 1872 936 kbps
4, with 3 2880 5760 5616 2.8 Mbps 2 Mbps
parallel
codes
• The number of orthogonal channelization codes = Spreading factor
151
maximum bit rate discontinuous transmission with lower bit rate
TS TS TS TS TS
Multicode usage:
DPCH 1
TS TS TS
DPCH 2
TS TS TS
DPCH 3
152
• Power offsets
Power Offsets for the DPCH • TFCS
• DL DPCH slot
format
• FDD DL TPC step
size
NBAP: RADIO LINK SETUP REQUEST
DCH Data Frame

Iub
Uu
Node B RNC
P0x: 0..6 dB
step size: 0.25
dB UE
PO2 TPC TFCI PO3

PO1 Pilot bits
bits bits
Data 1 bits (optional) Data 2 bits
153
Nokia Parameters Related to DPCHs
• RNC: PowerOffsetDLdpcchPilot
The parameter defines the power offset for the pilot symbols in relative to the data
symbols in dedicated downlink physical channel
[0 … 6 dB]; step size 0.25 dB; default: 3 dB for 12.2 kbps
• RNC: PowerOffsetDLdpcchTpc,
The parameter defines the power offset for the TPC symbols relative to the data
symbols in dedicated downlink physical channel
[0 … 6 dB]; step size 0.25 dB; default: 3 dB for 12.2 kbps
• RNC: PowerOffsetDLdpcchTfci,
The parameter defines the power offset for the TFCI symbols relative to the data
symbols in dedicated downlink physical channel.
[0 … 6 dB], step size 0.25 dB; default: 3 dB for 12.2 kbps
154
Uplink Dedicated Physical Channels
The uplink dedicated physical channel transmission, we identify two types of physical channels:
Dedicated Physical Control Channel (DPCCH),
which is always transmitted with spreading factor 256 (3840/256=15Ksps=15Kbps). Following
fields are defined on the DPCCH:
- Pilot bits for channel estimation. Their number can be 3, 4, 5, 6, 7 or 8.
- Transmitter Power Control (TPC), with either one or two bits
- Transport Format Combination Indicator (TFCI), which is optional, and a
- Feedback Indicator (FBI). Bits can be set for the closed loop mode transmit diversity and site
selection diversity transmission (SSDT)
6 different slot formats were specified for the DPCCH. Variations exist for the compressed mode.
Dedicated Physical Data Channel (DPDCH),
which is used for user data transfer (Data is not only User Traffic/Data but could be also
High layer signalling (“Measurement Reports”, “RRC messages”,L3 signalling) . Its
spreading factor ranges between 4 and 256. 7 different slot formats are defined, which are set by
the higher layers.
The DPCCH and DPDCH are combined by I/Q code multiplexing with each multiframe.
Multicode usage is possible. If applied, additonal DPDCH are added to the uplink transmission, but
no additional DPCCHs! The maximum number of DPDCH is 6; when more than one DPDCH is used
(Multicodes) they all use SF = 4.
The transmission itself is organised in 10 ms radio frames, which are divided into 15 timeslots. The
timeslot length is 2560 chips.
155
Uplink Dedicated Physical Channels
Superframe = 720 ms
Radio Frame Radio Frame Radio Frame Radio Frame

0 1 2 71
10 ms Frame
DPDCH Data 1 bits
TFCI bits TPC

DPCCH Pilot bits FBI bits
(optional) bits
• 6 different slot formats
• Compressed mode slot
format for changed SF Feedback Indicator for
• 7 different & changed puncturing • Closed loop mode transmit diversity, &
slot
• Site selection diversity transmission
formats 156
(SSDT)
Discontinuous Transmission and Power Offsets
Discontinuous transmission (DTX) is supported for the DCH both uplink and downlink. If DTX is
applied in the downlink – as it is done with speech – then 3000 bursts are generated in one
second. (1500 times the pilot sequence, 1500 times the TPC bits) This causes two problems:
• Inter-frequency interference, caused by the burst generation. At the Node B, the problem can
be overcome with exquisite filter equipment. This filter equipment is expensive and heavy.
Therefore it cannot be applied in the UE. The UE‘s solution is I/Q code multiplexing, with a
continuous transmission for the DPCCH. DPDCH changes can still occur, but they are limited
to the TTI period. The minimum TTI period is 10 ms. The same effects can be observed, then the
DPDCH data rate and with it its output power is changing.
• 3000 bursts causes audible interference with other equipment – just see for example GSM. By
reducing the changes to the TTI period, the audible interference is reduced, too.
Determination of the power difference between the DPCCH and DPDCH
I/Q code multiplexing is done in the uplink, i.e. the DPCCH and DPDCH are transmitted with different
codes (and possible with different spreading factors). Gain factors are specified: βc is the gain factor for
the DPCCH, while βd is the gain factor for the DPDCH. The gain factors may vary for each TFC. There are
two ways, how the UE may learn about the gain factors:
• The gain factors are signalled for each TFC.
If so, the nominal power relation Aj between the DPDCH and DPCCH is βd/βc.
• The gain factor is calculated based on reference TFCs.
157
Discontinuous Transmission and Power Offsets
DPDCH
DPDCH
DPDCH
DPCCH DPCCH DPCCH
TTI TTI TTI
UL DPDCH/DPCH Power Difference:

two methods to determine the gain factors:
• signalled for each TFCs
• calculation based on reference TFCs
DPDCH
βd
Nominal Power Relation Aj = =
βc DPCCH
158
Transmit Diversity – Closed Loop Mode
• Closed loop mode transmit diversity

– Used in DPCH and PDSCH to improve DL performance based on feedback
information from UE
– Channel coding, interleaving and spreading are done as in non-diversity mode.
The spread complex valued signal is fed to both TX antenna branches, and
weighted with antenna specific weight factors w1 and w2. NodeB adjusts the
phase of antenna 2 (as antenna 1 taken as reference), so as to maximize the
power received by UE
– The weight factors are determined by the UE, and signalled to the UTRAN access
point (=cell transceiver) using the D-bits of the FBI field of uplink DPCCH
– The calculation of weight factor is the key point of closed loop Tx diversity
– There are two modes with different calculation methods of weight factor:
1. Mode 1 uses phase adjustment；the dedicated pilot symbols of two antennas are
different (orthogonal)
2. Mode 2 uses phase/amplitude adjustment； the dedicated pilot symbols of two
antennas are the same
159
Initial Uplink DCH Transmission
When we look to the PRACH, we can see, that a preambles were used to avoid UEs to access
UTRAN with a too high initial transmission power. The same principle is applied for the DPCH.
After PRACH procedure the UE transmits between 0 to 7 radio frames only the DPCCH uplink
(the period is called DPCCH power control Preamble), before the DPDCH is code multiplexed.
The number of radio frames is set by the higher layers (RRC resp. the operator). Also for this
period of time, only DPCCH can be found in the downlink.
The UE can be also informed about a delay regarding RRC signalling – this is called SRB delay,
which can also last 0 to 7 radio frames. The SRB delay follows after the DPCCH preamble.
How to set the the transmission power of the first UL DPCCH preamble? Its power level is
DPCCH_Initial_power = – CPICH_RSCP + DPCCH_Power_offset
The DPCCH Power Offset is retrieved from RRC messages. It’s value ranges between –164 and
–6 dB (step size 2 dB). CPICH_RSCP is the received signal code power on the P-CPICH,
measured by the UE.
160
Initial Uplink DCH Transmission
DPCCH only DPCCH & DPDCH
reception
at UE
trans-
mission
at UE T0
0 to 7 frames for
power control preamble
DPCCH only, DPCCH & DPDCH
always based on PCA1 PCA based on RRC
DPCCH_Initial_power = – CPICH_RSCP + DPCCH_Power_offset 161
Radio frame timing and access slot timing of downlink physical channels
Primary
SCH
Secondary
SCH
Any CPICH
P-CCPCH Radio framewith (SFN modulo 2) = 0 Radio framewith (SFN modulo 2) = 1
k:th S-CCPCH τS-CCPCH,k
τPICH
PICH for k:th

S-CCPCH
AICH access #0 #1 #2 #3 #4 #5 #6 #7 #8 #9 #10 #11 #12 #13 #14

slots
Any PDSCH
n:th DPCH τDPCH,n
HS-SCCH Subframe Subframe Subframe Subframe Subframe

Subframes #0 #1 #2 #3 #4
10 ms 10 ms
162
Part VII
WCDMA Planning
163
Radio Network Planning Process
DEFINITION PLANNING and IMPLEMENTATION O&M
System Coverage Capacity Network

Dimensioning Planning and Optimisation Optimisation
Site Selection
Traffic distribution Survey
Requirements Path loss measurements
and strategy prediction Pilot Power
for coverage, Soft handover
quality and Coverage Statistical
capacity, optimisation performance
Blocking objectives
per service analysis
Cell isolation
optimisation
164
Planning issues
• Planning should meet current standards and demands and also comply with future
requirements.
• Uncertainty of future traffic growth and service needs.
• High bit rate services require knowledge of coverage and capacity enhancements
methods.
• Real constraints
– Coexistence and co-operation of 2G and 3G for old operators.
– Environmental constraints for new operators.
• Network planning depends not only on the coverage but also on load.
Objectives of Radio network planning

• Capacity:
– To support the subscriber traffic with sufficiently low blocking and delay.
• Coverage:
– To obtain the ability of the network ensure the availability of the service in the entire
service area.
• Quality:
– Linking the capacity and the coverage and still provide the required QoS.
• Costs:
– To enable an economical network implementation when the service is established and a
controlled network expansion during the life cycle of the network.
165
Planning methods
• Preparation phase
– Defining coverage and capacity objectives
– Selection of network planning strategies
– Initial design and operation parameters
• Initial dimensioning
– First and most rapid evaluation of the network elements count and capacity of
these elements
– Offered traffic estimation
– Joint capacity coverage estimation
• Detailed planning
– Detailed coverage capacity estimation
– Iterative coverage analysis
– Planning for codes and powers
• Optimization
– Setting the parameters
• Soft handover
• Power control
• Verification of the static simulator with the dynamic simulator
166
A strategy for dimensioning
• Plan for adequate load and number of sites.
• Enable optimized site selection.
• Avoid adding new sites too soon.
• Allow better utilization of spectrum.
• Recommended load factor 30- 70 %
167
Dimensioning process
168
Detailed Planning – Workflow of RNP Tool
169
Input data preparation
• Digital map.
- for coverage prediction.
- topological data (terrain), morphological data (terrain type), building location
and height.
- Resolution: urban areas 10-20m, rural areas 50-100m.
• Plan.
- logical concept combining various items.
digital map, map properties, target plan area, selected radio access
technology, input parameters, antenna models.
• Antenna editor.
- logical concept containing antenna radiation pattern, antenna gain, frequency
band.
• Propagation model editor.
- Different planning areas with different characteristics.
- For each area type many propagation models can be prepared.
- tuning based on field measurements.
• BTS types and site/cell templates
- Defaults for the network element parameters and ability to change it.
- Example BTS parameter template:
maximum number of wideband signal processors.
maximum number of channel units.
noise figure. 170
Tx/Rx diversity types. Company Confidential
Planning
• Importing sites.
– Utilization of 2G networks.
• Editing sites and cells.

– Adding and modifying sites manually.
• Defining service requirements and traffic modelling.

– Bit rate and bearer service type assigned to each service.
– For NRT need for average call size retransmission rate.
– Traffic forecast.
• Propagation model tuning.

– Matching the default propagation models to the measurements.
– Tuning functions per cell basis.
• Link loss calculation.

– The signal level at each location in the service area is evaluated, it depends on
• Network configuration (sites, cells, antennas). Propagation model. Calculation
area. Link loss parameters. Cable and indoor loss. Line-of-sigth settings.
Clutter type correction. Topographic corrections. Diffractions.
• Optimising dominance.
– Interference and capacity analysis.
– Locating best servers in each location in the service area.
– Target to have clear dominance areas. 171
Iterative traffic planning process
• Verification of the initial dimensioning.
• Because of the reuse 1, in the interference calculations also interference

from other cells should be taken into account.
• Analysis of one snapshot.

– For quickly finding the interference map of the service area.
– Locate users randomly into network.
– Assume power control and evaluate the SIR for all the users.
– Simple analysis with few iterations.
– Exhaustive study with all the parameters.
• Monte-Carlo simulation.
– Finding average over many snapshots: average, minimum, maximum, std.
– Averages over mobile locations.
– Iterations are described by:
• Number of iterations.
• Maximum calculation time.
• Mobile list generation.
• General calculation settings.
172
Example of WCDMA analysis
• Reporting:
– Raster plots from the selected area.
– Network element configuration and parameter setting.
– Various graphs and trends.
– Customized operator specific trends.
173
Uplink iteration step
• Allocate MS transmit powers so that
the interference levels and BS
sensitivities converge.
• Transmit power of MS should fulfill

required receiver Eb/Io in BS.
– Min Rx level in BS.
– Required Eb/Io in uplink.
– Interference situation.
– Antennae gain cable and other losses.
• The power calculation loop is

repeated until powers converge.
• Mobiles exceeding the limit power

– Attempt inter-frequency handover.
– Are put into outage.
• Best server in UL and DL is selected.
174
Downlink iteration step
• Allocation of P-CPICH powers.

• Transmit power of BS should fulfill required receiver Eb/Io in MS.
• The initial Tx powers are assigned iteratively.
• The planning tool evaluates the actual CIR and compares it to the Target CIR
175
Coverage Analysis
UL DCH Coverage
– Whether an additional mobile having certain bit rate could be served.
– The transmit power need for the MS is calculated and compared to the
maximum allowed
DL DCH Coverage
– Pixel by pixel is checked whether an additional mobile having certain bit
rate could be served. Concentration on the power limits per radio link.
– The transmit power need for supporting the link is calculated and
compared to the maximum allowed
DL CPICH Coverage
– Pixel by pixel is checked whether the P-CPICH channel can be listened.
176
Examples of Plots/Arrays
177
178
179
180
Capacity&Coverage Trade Off
• The coverage for a WCDMA system is generally limited by the uplink. This is
because the maximum output power of the mobile is lower than for the base
station, so the base station can reach longer than the mobile can.
• Capacity is generally limited by the downlink. This is because better receiver

techniques can be used in the base station than in the mobile. Since most
forecasts predict an asymmetric load where the users download data to a larger
extent than sending, the downlink will be most important from a capacity point of
view.
• Capacity and coverage is closely related in a WCDMA system. When traffic

increases, the level of interference in the system increases. To compensate for
this, the mobile has to increase its output power in order to defeat the increased
noise, or in already at max power, make the connection closer to the base station.
• Due to the increase of traffic, the effective cell area has shrunk. This behavior is
known as cell breathing.
breathing In an FDMA or TDMA-system this problem does not arise,
since coverage and capacity is largely independent.
• To reduce cell breathing interference margins are included when dimensioning the
network, which has the effect of increasing site density. 181
Coverage Limited Uplink
• Another way to reduce cell breathing would be to add a frequency, which would
mean that the users could be spread over two or more carriers. Since the different
carriers are not interfering with each other, the interference level is reduced, and
an increase in capacity or coverage is achieved
• When making the initial design, the aim is to provide a certain capacity, or service
level, over an area. One design strategy could be to design a very low-density
network, capable of providing low capacity over a wide area.
• This would reduce the number of base stations as compared to building for higher
capacity. Since the cost of base stations are a large part of the cost of building a
network, minimizing the number of base stations are important.
• On the other hand, it is important to be able to provide attractive services to the
customers. This could be difficult if not enough bandwidth is available. Building
less dense means that the maximum distance between the mobile and base
station is increased, which is the same as allowing a higher maximum path loss
between the two.
• A higher path loss between the mobile and the base station can be tolerated if the
interference is decreased. If the interference in a cell were reduced by a certain
amount of dB, the maximum allowed path loss would increase by the same
amount.
182
• Using a propagation model like for example Okumura-Hata, it is possible to
convert a change of the interference level into a changed site density, compared
to a reference case.
• Table below shows the change in number of sites if the interference margin in
the link budget is changed. A negative dB value means that the link budget is
worse compared to the reference case, and thus more sites are needed.
183
Uplink Load Factor
Interference degradation margin: describes the amount of increase of interference
due to multiple access . It is reserved in the link budget.
Can be calculated as the Noise Rise: the ratio of the total received power Itotal to
the Noise Power PN
I total 1 1
Noise Rise = = =
1 − ∑ j =1 L j 1 − ηUL
N
PN
Where Load Factor η ULis :
ηUL = ∑ j =1 L j , L j is the load factor of one connection
N
The uplink load factor can be written as

N 1
ηUL = (1 + i ) ⋅ ∑ j =1 L j = (1 + i ⋅ S ) ⋅ ∑ j =1
N N
ζ 1+
W
( Eb / N O ) j ⋅ R j ⋅ υ j
Noise Rise (dB) is equal to - 10 ⋅ log10 (1 − ηUL )
Where ηUL ∈ [0,1]
184
Uplink Load Factor
Definitions Recommended Values
N Number of users per cell
υj Activity Factor of user j at physical layer 0.67 for speech
Eb/No Signal energy per bit divided by noise spectral Dependent on service, bit rate, mulitpath,
density that is required to meet a predefined fading channel, receive antenna diversity,
BLER. Noise includes both thermal and mobile speed, etc
interference
W WCDMA chip rate 3.84 Mcps
Rj Bit Rate of user j Dependent on service
Sectorisation Gain 1 Sector (Omni): 1; 3 Sectors (90°): 2.57;

ζ 3 Sectors (65°): 2.87; 3 Sectors (33°): 2.82
4 Sectors (90°): 3.11; 6 Sectors (65°): 4.70
NS Number of Sectors
i Other cell to own cell interference ratio seen by Macro Cell with omni antennas: 55%. Macro
the base station receiver Cell with 3 sectors: 65% 185
Uplink Noise Rise as a Function of
Throughput
12
11
Noise Rise (dB)
10
9
8
7
6
5
4
3
2
1
200 400 600 800 1000 1200 1400 1600
Throughput (Kbps)
144 Kbps
Voice (12.2 Kbps)
186
• For voice services a typical noise rise would be between 1-3 dB, which
corresponds to a throughput between 150 kbps and 375 kbps. A network is
designed for a certain throughput.
• After some time that throughput is reached, and as a result the noise rise rises
over the design value.
• The choice is then to either increase site density, or add more frequencies.
Adding a frequency has its own set of problems, most notably that soft handover
does not work between frequencies. This problem is less of an issue if new
frequencies are added to a number of sites over a wider area.
• The mobile can then move freely on the frequency it has been assigned, and the
probability of making a hard inter-frequency handover is reduced
• Assume that traffic increases so that the actual noise rise is 4 dB, 1 dB above the
design level. The noise figure needs to be improved, for example down to 2 dB,
to improve quality and make room for future capacity demands. In other words,
the average throughput per cell needs to be reduced.
• Building more sites, or adding another frequency can do this.
• Adding a second frequency would half the throughput for each cell and carrier.
For a 4 dB noise rise the throughput is 450 kbps according to the graph. A new
throughput of 450/2 kbps=225 kbps per carrier gives a noise rise of 1.5 dB, an
improvement of 2.5 dB. A 2.5 dB lower allowable path loss corresponds roughly
to 40% more sites, that is. The cost of building these sites can then be said to be
the value of having one extra frequency. Adding a second and a third frequency
follows the same pattern, with a slight difference. The relative decrease in noise
rise will be lower. When a third frequency is added the traffic is spread over three
carriers, and reduced with a third for each frequency. 187
• It is also possible to do the other way around, that is, build sites less dense to
start with. This saves money in the roll out phase, but could cause problems if
high capacity is needed in the future. Using figures from the example above,
assume a design for a maximum throughput of 375 kbps for one carrier, which
corresponds to a noise rise of 3 dB. Using two carriers gives a throughput per
carrier of 375/2 kbps=190 kbps, which corresponds to a noise rise of 1.3 dB.
The saving is 2.7 dB, which converts to roughly 70% of the original number of
sites is needed. This is the same as each site covers approximately 1.4 times
the area of the original one carrier site.
Uplink Coverage of Different Bit Rates

3.5
2.5
Range [km]
1.5
0.5
0
32 kbps 64 kbps 144 kbps 384 kbps 1024 kbps 2048 kbps
Suburban area with 95% outdoor location probability 188

Downlink Load Factor
η DL = ∑ j =1υ j ⋅
N ( Eb / N O ) j
W Rj
[
⋅ (1 - α j ) + i j ]
Noise Rise over therm al noise due to multiple access interferen ce
is equal to - 10 ⋅ log 10 (1 − η DL )
Where η DL ∈ [0,1]
Definitions Recommended Values

N Number of users per cell
υj Activity Factor of user j at physical layer 0.58 for speech
Eb/No Signal energy per bit divided by noise spectral Dependent on service, bit rate, mulitpath,
density that is required to meet a predifined fading channel, receive antenna diversity,
BLER. Noise includes both thermal and mobile speed, etc
interference
W WCDMA chip rate 3.84 Mcps
Rj Bit Rate of user j Dependent on service
αj Orthogonality of channel of user j Dependent on the multipath propagation
1: fully orthogonal 1-path channel
0: no orthogonality
ITU Vehicular A channel: ~ 50%
ITU Pedestrian A channel: ~ 90%
ij
ij Ratio of other cell to own cell base station power, Each user sees a different , depending on
received by user j its location in the cell and log-normal
shadowing. Macro Cell with omni antennas: 189
55%. Macro Cell with 3 sectors: 65%
Downlink Load Factor
• Compared to the uplink load equation, the most important new parameter is α j ,
which represent the orthogonality factor in the downlink. WCDMA employs
orthogonal codes in DL to separate users, and without multipath propagation the
orthogonality remains when the base station signal is received by the mobile.
• The DL load factor exhibits very similar behavior to the UL load factor, in the
sense that when approaching unity, the system reaches its pole capacity and the
noise rise over thermal noise goes to infinity
• For downlink dimensioning, it’s important to estimate the total amount of base
station transmission power required. This is based on average transmission
power for user
(Eb No ) j
N rf ⋅W ⋅ L ⋅ ∑ j =1υ j
N
BS _ TxPw =
(W Rj )
1 − η DL
Where N rf is the noise spectral density of the mobile
N rf = k ⋅ T + NF = −174dBm + NF (assuming T = 290K)
k is the Boltzmann constant of 1.381⋅10-23 J / K , NF is mobile Noise Figure (5 - 9 dB)
190
Downlink Common Channels
• Part of the downlink power has to be allocated for the common channels that
are transmitted independently of the traffic channels
Downlink common Relative to CPICH Activity Average Power allocation
channels with 20W max Power
CPICH 0 dB 100% 2.0 W (33 dBm)
P-SCH -3 dB 10% 0.1 W
S-SCH -3 dB 10% 0.1 W
P-CCPCH -5 dB 90% 0.6 W
PICH -8 dB 96% 0.3 W

Almost
AICH -8 dB 80%¹ 0.25 W
50% is for
S-CCPCH 0 dB² 25%³ 0.5 W CPICH
Total Common 3.85 W

channels Power
Remaining power for 20-3.85 = 16.15 W
traffic channels
¹ Worst case; ² Depends on the FACH bit rate; ³ Depends on PCH and FACH traffic
191
Total DL Common Channel Power
Service Default Minimum Minimum Maximum Maximum
Type Power Activity Average Activity Average
Power Power
CPICH 33 dBm 100 % 33 dBm 100 % 33 dBm
P-SCH 30 dBm 10 % 20 dBm 10 % 20 dBm
S-SCH 30 dBm 10 % 20 dBm 10 % 20 dBm
P-CCPCH 28 dBm 90 % 27.5 dBm 90 % 27.5 dBm
S-CCPCH 33 dBm 25 % 27 dBm 115 %* 33.6 dBm
PICH 25 dBm 96 % 24.8 dBm 96 % 24.8 dBm
AICH 25 dBm 0% - 80 % 25 dBm
Total - - 35.5 dBm - 37.5 dBm
3.5 W 5.6 W
* S-CCPCH control (TFCI) bits transmitted with higher power than data bits
192
Relation of Uplink and Downlink Load
100
90
• Downlink load is always
80 higher than uplink load due to:
70
– asymmetry in user traffic
DL Load [%]
60
– different Eb/No values in
50 Increasing uplink and downlink
asymmetry
40
– orthogonality in downlink
30
– overhead due to soft-
20 handover
10
0
0 10 20 30 40 50
UL Load [%]
193
Capacity Limited Downlink
• As the demand for downlink capacity increases, there are several different ways of
increasing capacity. The most common ways are adding more frequencies and
power amplifiers, and introducing transmit diversity
• Upgrading capacity in the ways just mentioned is of course dependant on the base
station equipment being able to handle it. It is reasonable to assume that as the
capacity demand increases, the equipment vendors will produce equipment that
can handle it
• Assume an initial base station configuration of one 20W power amplifier per sector,
one carrier per sector and three sectors per site. This is called the baseline
configuration, and has a baseline capacity
• The first step to upgrade the capacity is to add a second frequency. This gives a
capacity increase of 80%. The reason why the increase is not 100% is that the
power amplifier only can deliver 20W, which has to be split between the two
frequencies, making the output 10W per carrier.
• The second step could be to add a second 20W power amplifier (restoring the
power per frequency to 20W) and introduce transmit diversity (STTD, Closed loop
mode 1, Closed loop mode 2).
• With these two upgrades the capacity now is 180% compared to the baseline.
Adding a third frequency would decrease the output power to 13 W per carrier, but
the extra carrier would still mean a capacity increase of 290% compared to the
baseline
• If there are no more frequencies available, changing the power amplifiers from two
20W to two 40W will give a modest capacity increase, making the increase
compared to baseline 320%. Adding a fourth frequency and at the same time
changing out the two 20W power amplifiers to two 40W amplifiers, if that has not
194
been done before, gives a capacity increase 460% compared to the baseline.
Capacity Limited Downlink
• Upgrading the power amplifier restores the power per frequency to 20W, the
same as the baseline case. With the stronger PA’s there is power to add a fifth
and a sixth carrier. This would give capacities of 550% respectively 680%
compared to the baseline
• Using two PA’s means that no modification to the antenna system is required.
Adding a third PA means that either a combiner, or an extra antenna needs to be
used. A combiner typically has a 3dB insertion loss, offsetting the gain achieved
• Adding a third antenna is complicated from a site-engineering point of view. An
extra feeder cable is needed, and adding an extra antenna could be difficult since
it means renegotiating the agreement with the house owner.
• With a third PA the 6 frequencies is transmitting at 20W, giving a 740% increase
gain compared to the baseline capacity.
Example upgrade path 195

Typical Pathlosses for different Bearer Services
Low Data Scenario

Low Asymmetry Scenario
165,00
160,00
Speech 12,2k UL Pathloss

155,00
Pathloss [dB]
RT Data 14k UL Pathloss

RT Data 64k UL Pathloss
better coverage
NRT Data 144k UL Pathloss

NRT Data 384k UL Pathloss
150,00
DL Pathloss
145,00
Coverage is
uplink limited Capacity is
downlink limited
140,00
0 10 20 30 40 50 60 70 80
UL Load
196
Part VIII
WCDMA Link Budget
197
WCDMA Link Budget
198
WCDMA Link Budget
199
WCDMA Link Budget
200
WCDMA Link Budget
201
WCDMA Link Budget
202
WCDMA Link Budget
203
WCDMA Link Budget
204
WCDMA Link Budget
205
WCDMA Link Budget
206
WCDMA Link Budget
207
WCDMA Link Budget
208
WCDMA Link Budget
209
WCDMA Link Budget
210
WCDMA Link Budget
211
WCDMA Link Budget – Cell Sizes
•Output of Link Budget is MAPL (Maximum Allowed Path Loss) based on different:
- Clutter types (Dense Urban, Urban, Sub-Urban, Rural)
- Services (AMR, PS64, CS64, PS128, PS384,…)
- Indoor/Outdoor
- Area Location Probability
- Mobile speed: Pedestrian/Vehicular
•Given an area to be covered (Km²) the cell count has to be performed based on
Cell Area ( Cell Radius)
•Cell Radius can be calculated using Propagation models (Cost231, Hokumura-
Hata, Walfish-Ikegami,…)
•The models need as input:
•MAPL
•UE antenna height
•NodeB antenna height
•Frequency
•Clutter correction factors
212
WCDMA Link Budget – Cell Count
r = Cell Radius
r
3 2
Surface of a tri-sectorial cell : ACell =3 r
2
Number of Sites = Number of Cells /3
Intersite Distance = 1.5*r

Example:
r = 0.409 km
Acell = 0.432 km2
Stotal = 100 km²

Number of Tri-secotrial sites = 100/0.432 ≈ 230
213
Part IX
Coverage-Capacity
enhancement
214
Coverage Improvement Alternatives
• Mast head amplifier
– basic solution for optimized uplink performance
– compensates feeder cable loss
– supported by Nokia's base stations
– can be used together with Smart Radio Concept
• 6 sectored site
– utilizing narrowbeam antennas
– ~ 2 dB better antenna gain than in 3
sectored site
• Nokia Smart Radio Concept, SRC

– 4-branch uplink diversity
215
Capacity Improvement Alternatives
• 6 sectored site
– ~ 80% capacity gain compared to 3
sectors (not 100% due to inter-sector
interference)
• More carriers (frequencies) per sector

– doubling the amount of carriers with
power splitting gives roughly 60% more
capacity
• Smart Radio Concept

– transmit diversity
216
Smart Radio Concept
Uplink coverage
– 4-branch diversity reception per sector dB Received signal power
10
– Maximal ratio baseband combining of 4
uplink signals forms a beam 5
Downlink capacity upgrade -5

– Upgrade transmit diversity when needed -10 SRC
Rx diversity
-15
0 0.5 1 1.5 2 2.5
+ TX
RX Seconds, 3km/h
RX
RX
RX + TX
WCDMA Combined
Transceiver received
signal
217
144 kbps Coverage / Capacity in Macro Cells
Max. allowed
path loss [dB]
170
Downlink
load curve
165
Better 160
coverage Coverage is
uplink limited
155 Capacity is
downlink limited
150 Uplink load
curve with RX
diversity for
144 kbps
145
100 200 300 400 500 600 700 800 900 1000110012001300
Load per sector [kbps]
218
Nokia Smart Radio Concept
Phase 1: Increase Uplink Coverage
Max. allowed
path loss [dB]
170
165
Uplink 2.5-3.0 dB
load curve coverage
160 with SRC improvement
with SRC
155
150 Uplink load

curve
without
145 SRC
100 200 300 400 500 600 700 800 900 1000110012001300
219
Nokia Smart Radio Concept
Phase 2: Increase Downlink Capacity
Max. allowed
path loss [dB]
170
Downlink with TX
diversity, 20W per
165 branch
160 Downlink
20W
no diversity
155
150
70% increase
in capacity
145
100 200 300 400 500 600 700 800 900 1000110012001300
220
Coverage : 30 % less sites with SRC
Sites / km2
0.3
0.25
2.5 - 3.0 dB gain
0.2 corresponds to 30%
less sites with SRC
0.15
0.1
0.05
0
3-sector (rx div) 3-sector (SRC)
221
Capacity Upgrade with Smart Radio Concept
• No changes to antennas or antenna cables
• All these capacity upgrades within one Ultrasite cabinet
Speech Erlang per site

350
300 Cost / Erlang is
decreasing with
250
capacity upgrades
200
Add tx diversity +
150
take 2nd frequency
100 into use
50
0
20W 2x10W + 2x10W
Downlink power per sector 222
Capacity Optimisation
The impact of MHA, SRC & 6 -sector site
3G Radio Network Planning case study
Assumptions:
• The geographic area under study is defined by the suburban area of
London
• The site's location was given, antennas´ directions are the same as the
DCS network. Two antenna type has been used, 60 and 90 degrees
horizontal opening
• 1800MHz measurements provided. Assumption that narrow band
1800MHz propagation is representative of wideband 2GHz propagation
• 15dB of building penetration loss included in the link budget for Indoor
Coverage.
• Multiple simulation runs. MS positions and slow fading changed for each
run
223
Area under Investigation
Suburban area of London

12km by 11km
Intended to be representative of
suburban areas across the UK
Morphology analysis
Morphology Percentage Area
Suburban 65.2%
Open/Fields 13.7%
Open in Urban 10.1%
Industrial 6.3%
Roads in Urban 2.2%
Forest 2.0%
Urban 0.5%
Water 0.1%
224
Radio Network Configuration
51 sites (3 sector)
existing 1G and 2G sites, plus

sites to be acquired prior to 3G
Link level simulations used to define Eb/No

requirements, SHO Gain
Vehicular A channel assumed
Parameter Assumptions
Parameter Value
Max. transmit power 43dBm
Max. power per link not limited
Min. transmit power per link not limited
CPICH power 30dBm
Common channel power 30dBm
Cable/connector loss 3dB
Soft handover window 5dB
RF carriers available 1
Slow fading standard dev. 8dB
Maximum uplink load 50%
225
Antenna Configuration
1G and 2G antenna list 3G antenna list
60° antenna x 56 60° antenna x 56
85° antenna x 97 90° antenna x 97
Differences in elec. tilt compensated with mech. tilt

Antenna Type Horizontal Vertical Electrical Antenna Gain
Beamwidth Beamwidth Downtilt
741415 60° 7° 0° 18dBi
CS72138 90° 7° 2° 16.5dBi
741415 CS72138
226
Traffic Modeling
Priority placed on modeling Uniform distribution of mobile terminals
traffic services separately
System loaded to maximum capability
12.2kbps speech
fixed uplink load limit
64kbps data fixed BTS power capability
144kbps data
Link level simulations used to define Eb/No
Symetric data services requirements, SHO Gain
Service
Parameter 12.2 kbps 64 kbps data 144 kbps data
voice MS Numbers
Max. transmit power 21 dBm 21 dBm 21 dBm
Min. transmit power -50 dBm -50 dBm -50 dBm
Distrib. Supp.
Antenna height 1.5 m 1.5 m 1.5 m
Antenna gain 0 dBi 0 dBi 2 dBi 12.2kbps speech 15000 ~5400
Body loss 3 dB 0 dB 0 dB
Uplink bit rate 12.2 kbps 64 kbps 144 kbps
64kbps data 5000 ~1100
Downlink bit rate 12.2 kbps 64 kbps 144 kbps
Uplink activity factor 0.67 1 1
144kbps data 1500 ~500
Downlink activity factor 0.67 1 1
Mobile speed 50 km/hr 3 km/hr 3 km/hr
227
Benchmark Results
MHA, SRC, 6 Sector not included
Only coverage and capacity presented Uplink limited

here
Capacities
Service Envir. network per cell
12.2kbps Outdoor 5074 33.2
Speech Indoor 5336 34.9
64kbps Outdoor 966 6.3
Data Indoor 1100 7.2
144kbps Outdoor 470 3.1
Uplink Data Indoor 501 3.3
limited
12
Depends 100
Percentage of Cells
Number of Cells
upon
polygon
6
Envir. Service of the Mean
Probe Mobile
Outdoor Speech 99.83
64kbps Data 98.54
144bps Data 96.74
Indoor Speech 88.05 0 0
0 30 60
64kbps Data 70.05
144bps Data 59.71 Number of 12.2kbps speech users
228
Impact of MHA
MHA introduced at all sites
Improves uplink power budget
Improved
Example
indoor speech: 88 93%
indoor 64kbps data: 70 79%
Trend of results as Remains approximately the

expected same
uplink and downlink
approximately balanced
229
Impact of SRC (rx only)
2 scenarios
SRC introduced at all sites
SRC introduced at TACS sites only
Reduces uplink Eb/No target
Improved
speech by 9%
64kbps data by 11% Improved
144kbps data by 30%
Becomes limited by BTS tx power indoor speech: 88 92%
Observations:
Remains uplink limited
Once downlink limited, soft handover window has
great impact upon capacity
Introducing SRC at TACS sites only, increases

capacity of surrounding sites
230
Impact of SRC (rx & tx)
Tx and Rx SRC introduced at all sites

Reduces uplink and downlink Eb/No
targets
Further Improved
Example: outdoor speech

Tx SRC offers no
no SRC 5000 users
coverage improvement over Rx SRC
rx SRC 5800 users (+15%)
rx&tx SRC 7500 users (+49%) Coverage remains uplink limited
231
Impact of 6 Sectors
2 scenarios
6 Sector introduced at all sites
6 Sector introduced at TACS sites only
33º beam width antennas increased gain
Doubled
Usual message for 6S:
Example, outdoor
capacity improves by
speech, 2500 5100 users less than a factor of 2
64kbps data, 440 950 users due to increased SHO
144kbps data, 230 450 users & inter-cell
interference
Improved
Example,
Not true in this case due to antenna
indoor speech: 85 93% pattern selected:
indoor 64kbps data: 65 80 %
reduced SHO and inter-cell
indoor 144kbps data: 50 70 % 232
interference
Part X
WCDMA/GSM
Co-siting issues
233
Antenna System Co-siting
• GSM 900 / GSM 1800 shared

antenna lines by
diplexers/triplexers
• GSM 900 / GSM 1800/WCDMA
multi band antennas
234
Antennas: WCDMA/GSM Co-site
Antenna 1: Antenna 2:
3 sector site:
Shared antenna lines GSM 900 / 1800 WCDMA X-pol
GSM 900/1800 antennas: 3 pcs
Dual Band X-pol
• GSM 900 / GSM 1800 / WCDMA WCDMA antennas:
WCDMA MHAs:
3 pcs
6 pcs
triplexers Triplexers: 6 pcs
Feederlines: 6 pcs
Shared antennas WCDMA WCDMA
• Dual Band GSM 900 / WCDMA MHA MHA
• Dual Band GSM 1800/WCDMA

• Triple Band 900/1800/WCDMA GSM 900 / GSM 1800 / WCDMA Triplexer-2
Mast Head Amplifiers

• Triplexer supports MHA in one GSM 900 / GSM 1800 / WCDMA Triplexer-1
branch
• Additional MHAs to be equipped
with direct DC feed
GSM 900 GSM 1800 WCDMA
BTS BTS BTS with Bias-Ts
235
Upgrades to Current GSM Antennas
Upgrade :
Current : space +
space polarization
diversity diversity
Space
Space diversity
diversity
improves
improves performance
performance
0.5..1.0
0.5..1.0 dB
dB compared
compared
to single radome.
to single radome.
The
The gain
gain of
of 2.5
2.5 dB
dB
assumes
assumes single
single radome.
radome. Antennas can
Current : be shared
1300 mm
polarization Upgrade: with GSM

diversity 2 x polarization
diversity within
one radome
236
150 mm 300 mm
Example: common feeders, separate antennas
• GSM 900/1800 BTS & WCDMA BTS
• Triplexers
– common feeders
• Separate antennas
– 900/1800 MHz dual-band
– 2 GHz
DPX
TPX
DPX
Triplexer
Abis/Iub
To/From
BSC/RNC Triplexer
Iub
Power
Site GSM WCDMA

Support BTS
System BTS
237
Nokia Base Stations and Co-Siting
Air-interface issues
• WCDMA - WCDMA Co-Siting

– This has been taken into account in 3GPP Air Interface Specifications
– Nokia WCDMA base station products are compliant with 3GPP
• WCDMA - GSM900 Co-Siting
– This has been taken into account with Nokia's WCDMA and GSM900
base station design
• WCDMA - GSM1800 Co-Siting
– This is as with GSM900
– If GSM1800 Transmitter Frequency separation within same sector is
more than 57 MHz( bottom channels) or 40 MHz (top channels), extra
transmitter filtering (~10 dB) may be required in GSM1800 BTS
• Note: 30 dB Minimum Coupling Loss (MCL) assumed between antennas
238
Co-Siting with other manufacturers
Air-interface issues
• WCDMA Co-Siting with other manufacturers'
equipment
– theoretical worst case requires 50 dB extra
isolation in GSM BTS
– in practice this much will not be needed
– Nokia can provide assistance with co-siting
issues
• Note: 30 dB Minimum Coupling Loss (MCL)
assumed between antennas
239
WCDMA - GSM Interference Outline
• Spurious emissions
• Nonlinear distortion
• Specifications and isolation requirements
• Interference mitigation methods
• Co-located sites
240
WCDMA - GSM Interference Outline
• Site and equipment sharing is an important issue to cut costs down
and to guarantee proper function of the networks.
• Common
• base station mechanics
• site support
• transmission
• antennas and feeders
• site construction
• network management
• By proper site design (antenna installation etc.) interference coupling
between systems can be reduced and unreasonable degradation of
service due to co-sited installations avoided.
• Co-siting preferred to avoid high path loss differences between own
and neighbour systems.
241
Spurious emissions
• ITU-R definition of Spurious Emission (ITU-R: 329-7_ww7.doc):

– Spurious Emission: Emission on a frequency or frequencies which
are outside the necessary bandwidth and the level of which may be
reduced without affecting the corresponding transmissions of
information. Spurious emissions include harmonic emissions,
parasitic emissions, intermodulation products and frequency
conversion products, but exclude out-of-band emissions.
• Normally the intermodulation distortion (IMD) is handled separately
due to its importance.
• Spurious signals can be coupled by
– radiation
– conduction
– combination of radiation and conduction
242
Nonlinear system
• Nonlinear system transfer function can be expressed as a series expansion
x System y = a0 + a1x + a2x2 + a3x3 + ...

• In the case of one input frequency, vin = cos ω1t, output will consist of
harmonics, mω1
– Fundamental (m = 1) frequency is the desired one.
– If m > 1, there are higher order harmonics in output => harmonic
distortion.
– Can be generated both inside an offender or a victim system.
• In the case of two input frequencies, vin = cos ω1t + cos ω2t , output will
consist of harmonics mω1 + nω2, where n and m are positive or negative
integers.
– Intermodulation is a process generating an output signal containing
frequency components not present in the input signal and it is called
intermodulation distortion (IMD).
– Most harmful are 3rd order (|m| + |n| = 3) products.
– Can be generated both inside an offender or a victim system.
243
Nonlinear components
• Nonlinearities of active components like amplifiers under normal
operation.
• Nonlinearities of passive components
– Antennas
– Feeders
– Connectors
• Antenna mismatching
– Reflected wave can cause IMD in the power amplifier.
• Damaged feeders => mismatching
• Loose connectors => mismatching, reflections and rectification.
244
Active nonlinear distortion
• Active nonlinear distortion is generated in nonlinearities of active components
like amplifiers and modulators
• The nonlinearity effect is especially strong in power amplifiers if they
are driven to saturation.
• Intermodulation levels of the amplifiers can be decreased by
backing-off of them.
3rd order intercept
• The amplitude of the 3rd order product point
increases 3 dB compared to the fundamental
frequencies due to x3 term of it.
Desired signal
• Active IMD generated inside an offender BTS slope = 1
can be removed by BTS TX filtering.
3rd order
IMD
slope = 3
245
Passive nonlinear distortion
• Passive nonlinear distortion is generated in

nonlinearities of passive components like I
connectors, antennas and feeders.
• Contact and material nonlinearities
– Loose connectors
V
– Oxidation of joints
– Cracks in materials
– Electron tunneling through layers
– Nonlinear resistivity of materials
B
– B/H nonlinear hysteresis
• Levels normally lower than in active IMD.
• Aging of the components increases IMD
• Can NOT be filtered out in BTS TX.
H
246
Harmonic distortion
• Harmonic distortion can be a problem in the case of co-siting of

GSM900 and WCDMA.
• GSM900 DL frequencies are 935 - 960 MHz and second harmonics
may fall into the WCDMA TDD band and into the lower end of the
FDD band.
2nd harmonics
• 2nd harmonics
can be filtered
out at the
fGSM = 950 - 960 MHz output of
... GSM900 BTS.
GSM900 WCDMA WCDMA FDD

935 - 960 MHz TDD 1920 - 1980
1900 -
247
1920 MHz
IMD3 from GSM1800 DL to WCDMA UL
• GSM1800 IM3 products are hitting
into the WCDMA FDD UL RX band if • For active elements IM
products levels are higher
• 1862.6 ≤ f2 ≤ 1879.8 MHz than IM products produced
• 1805.2 ≤ f1 ≤ 1839.6 MHz by passive components
• Typical IM3 suppression
fIM3 = 2f2 - f1 values for power amplifiers
are -30 … -50 dBc
depending on frequency
spacing and offset
• Typical values for passive
f1 f2 elements are
-100 … -160 dBc
X dBc fIM3
GSM1800 GSM1800 WCDMA WCDMA

UL DL UL DL
1710 - 1785 MHz 1805 - 1880 MHz 40 MHz 1920 - 1980 MHz 2110 - 2170 MHz
248
Nonlinear distortion conclusions
• Second harmonics from the GSM900 system may fall into the WCDMA TDD
band.
• Intermodulation can be a problem if an operator has a splitted GSM1800 band or
in multioperator systems.
• The most harmful intermodulation products are 3rd order products which may fall
into the WCDMA RX band:
fIM3 = 2f1,2 — f2,1
• IM products can be avoided by proper frequency planning in GSM.
• fIM3 is hitting into the WCDMA FDD RX band (1920 - 1980 MHz) if GSM1800
channels are from 512 to 684 (f2) and from 799 to 885 (f1).
• Active intermodulation products can be filtered out in GSM1800 BTS TX
– IM products generated inside a WCDMA receiver cannot be filtered out.
• Passive IM products can not be filtered out in BTS TX if they are generated in
feeder lines and connectors after the filtering unit of BTS.
• Some aging problems may be avoided by installation, site administration and
maintenance recommendations.
249
RF Specifications
• GSM 05.05-8.7.1, WCDMA TS 25.104-3.5.0
• Two main reasons to isolate GSM and WCDMA
– Blocking
– Sensitivity
Transmitter Frequency Level Parameter affected Required Required

[MHz] [dBm] / [MHz] [dBm] / MHz isolation [dB]
GSM 1920 – 1980 −96 / 0.1 UMTS BTS < −108 / 4.0 28
spurious (FDD UL) -80 / 4.0 sensitivity (Noise floor)
GSM 1805 − +40 / 0.2 UMTS BTS < −15 / CW 55
main 1880 Typical blocking (Specifications)
UMTS 1710 − −98 / 0.1 GSM BTS < −110 / 0.2 15
spurious 1785 −95 / 0.2 sensitivity (Typical)
UMTS 2110 – 2170 +43 / 4.0 GSM BTS 0 43
main (FDD DL) Typical blocking (Specifications)
250
Interference mitigation methods
• Means to achieve the required isolation

– RF-methods
• Tighter filtering of the GSM BTS TX signal
• Proper frequency planning in GSM
• Di- or triplexer in case of feeder and antenna sharing between
different systems
• By proper antenna selection and placing
– Baseband methods
• Interference cancellation receivers
• If the interferer is known its effect can be removed easily
– Combined RF and baseband methods
251
Antenna isolation measurements
• Measurements performed in an anechoic room in a GSM1800 band with a

HP8753/D network analyzer.
• According to the most common definition the far field assumption is valid if
2D 2
df = , and d f , D >> λ
λ
where D is the largest dimension of an antenna, λ is wavelength and df
is the distance from antenna.
• The far field assumption is not valid => measurements needed.
• For a typical GSM1800 antenna dimensions (D ≈ 1 m) df ≈ 13 m.
• Let's assume coupling loss of 65 dB from the near field to the far field =>
– Extra 10 dB means therefore about 30 m distance by deploying a free
space model from d0 = 10 m.
252
Isolation measurements
Antennas and configurations
Horizontal
Antenna beamwidth Gain Polarisation Frequency band
Vert. Pol A 65º 18 dBi Vertically linear 1710 – 1880 MHz
B 90º 16 dBi Vertically linear 1710 – 1880 MHz
C 90º 17.5 dBi Vertically linear 1710 – 1880 MHz
Dual. Pol D 90º 16 dBi +/- 45º dual pol. 1710 – 1880 MHz
120°
d d d d
d
I (90°)
II (120°) III (180°) IV (Horizontal) V (Vertical)
1TSG-RAN Working Group 4 (Radio) Meeting #8
TSGR4#8(99)631
Sophia Antipolis, France 253
26-29 October 1999
Source: Allgon Company Confidential
Setup Antenna d [mm] / Min d [mm] / Max
isolation [dB] isolation [dB]
I A 250 / 50 850 / 63
d B 250 / 46 975 / 59
d
C 250 / 54 950 / 62
D, Co-polar 200 / 46 1250 / 59
I (90°) D, Cross-polar 200 / 49 1000 / 58
II A Same mast / 49 1050 / 66
d
120° B Same mast / 38 1100 / 66
C Same mast / 53 1150 / 68
D, Co-polar Same mast / 38 1100 / 65
II (120°) D, Cross-polar Same mast / 43 1050 / 63
254
III A Same mast / 52 750 / 71
B Same mast / 49 1300 / 69
d
C Same mast / 52 1150 / 76
D, Co-polar Same mast / 38 1250 / 62
III (180°) D, Cross-polar Same mast / 53 1250 / 62
IV A 250 / 37 6000 / 57
B 250 / 27 6000 / 52
C 250 / 34 6000 / 48
D, Co-polar 250 / 33 4250 / 53
d
D, Cross-polar 250 / 36 6000 / 57
IV (Horizontal)
255
V A 2250 / 50 6000 / 70
d B 2250 / 55 5500 / 69
C 2250 / 61 6000 / 66
D, Co-polar 1500 / 42 6000 / 61
V (Vertical) D, Cross-polar 1500 / 44 5500 / 65
256
• Measurements performed in a more realistic environment by Nokia.
• The used antennas are listed in the table below
Band Manufacturer Model No Horizontal Polarisation Vertical Gain Electrical

Beamwidth Beamwidth Downtilt
UMTS Racal UMTSXP/65/17.7/2 65 deg. X-polar 7 deg 17.7dB 2 deg
GSM1800 CSA PCNV065-13-0B 65 deg. X-polar 7 deg 18 dBi 0 deg
GSM1800 CSA PCNA115-19-0B 115 deg. Vertical 5 deg 17dBi 0 deg
• Horizontal, vertical and combined displacement configurations measured.

• Rooftop, face and tower mounted measurements.
• Both co- and cross-polar feed used.
257
• Measured frequencies from 1710 to 1980 MHz and results collected from
1900, 1950 and 1980 MHz.
• Measurement corresponds spurious emissions attenuation from the
GSM1800 band into the WCDMA band.
output input
Antenna A (fixed) Network Analyser Antenna B
Figure 3. Equipment set up

258
Antenna isolation measurements: Horizontal
Antenna A Antenna B
(fixed) UMTS Front View
horizontal
separation
distance
Side View
direction of radiation
1000mm
2000mm
400mm 650mm
Figure 5. Sketch of measurement configuration 259

GSM1800 65 deg to UMTS 65 deg

Horizontal co-polar measurements
75.00
70.00
65.00
1900MHz
Isolation (dB)
1950MHz
60.00
1980MHz
55.00
50dB marker
50.00
45.00
40.00
00 00 00 00 00 00 00 00 00 00 ..
0. 1. 2. 3. 4. 5. 6. 7. 8. 9. 1.
Distance (m)
260

Horizontal co-polar measurements
65.00
60.00
55.00
Isolation (dB)
50.00
50dB marker
1900MHz
45.00
1950MHz
1980MHz
40.00
35.00
30.00
0
00
00
00
00
00
00
00
00
00
00
.0
6.
0.
1.
8.
2.
3.
4.
5.
7.
9.
10
Distance (m)
261

Horizontal measurements
60.00
55.00
Isolation (dB)
50dB marker
50.00 1900MHz
45.00 1950MHz
40.00 1980MHz
35.00
30.00
0
00
00
00
00
00
00
00
00
00
00
.0
.0
.0
0.
1.
2.
3.
4.
5.
6.
7.
8.
9.
10
11
12
Distance (m)
262
Antenna isolation measurements: Face
Antenna A
GSM1800 Front View
1m 5m
Antenna B
UMTS
Side View
1000mm
2000mm
300mm
400mm 650mm
Figure 9. Sketch of measurement configuration

263
Antenna isolation measurements: Face
Face mounting GSM1800 85 deg to UMTS 65 deg -

Co-polar
85.00
1900MHz
1950MHz
80.00 1980MHz
75.00
70.00
0.00 1.00 2.00 3.00 4.00 5.00
264
Antenna isolation measurements: Vertical
Antenna B
UMTS
Antenna A
GSM1800
(fixed)
10m
Figure 11. Sketch of measurement configuration 265

Antenna isolation measurements: Vertical

85.00 Noise Floor
Noise Floor
80.00
75.00
Isolation (dB)
70.00 1900MHz
1950MHz
65.00
1980MHz
60.00
55.00
50.00
0
25
50
25
50
0
0
0.
0.
1.
1.
0.
0.
1.
Distance (m)
266
Antenna measurement conclusions
• According to the measurements it's easy to find a configuration,

which provides isolation of 30 - 60 dB.
• Lowest isolation (27 dB) was measured in an anechoic room
antennas horizontally displaced 0.25 m
– with 6 m distance isolation was already about 50 - 55 dB.
• Highest isolation values were measured with the face mounted
antenna and the isolation was more than 70 dB.
• In Allgon's measurements both antennas were for GSM1800 and in
Nokia's measurements for GSM1800 and WCDMA.
– There is also attenuation between GSM1800 and WCDMA due to
frequency difference of them => isolation figures are higher for
the spurious emissions.
267
Isolation 800/900 - UMTS
Horizontal Separation : XPol 900 65° _ XPol UMTS
(824-960) (1710-2170)
268
(824-960) (1710-2170)
269
Vertical Separation : XPol 900 65° _ XPol UMTS
(824-960) (1710-2170)
270
(824-960) (1710-2170)
271
Separation by 120° : XPol 900 65° _ XPol UMTS
(824-960) (1710-2170)
272
(824-960) (1710-2170)
273
(1710-1990) (1710-2170)
274
(1710-1880) (1710-2170)
275
(1710-1990) (1710-2170)
276
(1710-1880) (1710-2170)
277
(1710-1990) (1710-2170)
278
(1710-1880) (1710-2170)
279
Isolation Dualband GSM 900/1800 - UMTS
Horizontal Separation : XXPol 900/1800 65°/65° _ XPol UMTS
(870-960/1710-1880) (1710-2170)
280
Isolation Dualband GSM 900/1800 - UMTS
Vertical Separation : XXPol 900/1800 65°/65° _ XPol UMTS
(870-960/1710-1880) (1710-2170)
281
Isolation UMTS - UMTS
Horizontal Separation : XPol UMTS 65° _ XPol UMTS
(1710-2170) (1710-2170)
282
Vertical Separation : XPol UMTS 65° _ XPol UMTS
(1710-2170) (1710-2170)
283
Separation by 120°: XPol UMTS 65° _ XPol UMTS
(1710-2170) (1710-2170)
284
Part XI
WCDMA Optimization
285
Network Optimization Process
Objective: To optimize the outdoor part of the 3G network, this done cluster wise,
as they are being integrated.
The main elements for this process are:

1.Pre-optimisationsurvey
2.Network check
3.Initial drive test, baseline
4.Pre-Launch optimization
•Cluster tuning until break-out point is reached
•Ready for network acceptance & reporting
286
Pre Launch Optimization-Overview
287
Pre Launch Optimization-Process
288
Optimization-Overview
289
Optimization-Overview Block A
290
Optimization-Overview Block A
291
Optimization-Overview Block B
292
Optimization-Overview Block C
293
Optimisation - required performance
• Examples of performance metrics

– Area of service availability or coverage performance
– Average FER, BLER
– Access failures including paging and SMS
– MOC/MOT Call Setup Failures
– Dropped call performance
– Handover percentage (Soft/Hard)
– Ec/Io&RSCP performance
• UMTS Bearer Service Attributes
– Maximum/Average bitrate (kbps)
– Residual bit error ratio
– Transfer Delay
– Guaranteed bitrate (kbps)
294
Key Performance Indicators, KPI
• KPIs are a set of selected indicators which are used for measuring the
current network performance and trends.
• KPIs highlight the key factors of network monitoring and warn in time of
potential problems. KPIs are also used to prioritise the corrective actions.
• KPIs can be defined for circuit switched and packet switched traffic
separately and be measured by field measurement systems and Nokia
NetActTM network management system.
• An example set of KPIs
– RRC Setup Complete Ratio
– RAB Setup Complete Ratio
– RAB Active Complete Ratio
– Call Setup Success Ratio
– Call Drop Rate
– Softer/Soft Handover Fail Ratio
295
WCDMA RAN Optimisation
Network Management WCDMA RAN
• Nokia NetActTM for 3G
• Field Tool Server configuration
KPIs, counters
me
as ur e
Configuration KPIs, me air-interface
measurements
nt s
RAN Optimisation
• pre-defined procedures
• semi / full automated Start
WindowAdd WindrowDrop CompThreshold DropTimer

Change 1 stepsize Change 1 stepsize Change 1 stepsize Change 1 stepsize
NMS: Collect
network
performance data
No
Evaluate KPI
'HO Overhead'.
OK ?
Field Tool
Yes
Evaluate all Go to relevant

network KPIs. No optimisation
OK ? flow-chart
Yes
End
296
WCDMA Field Tool
• Measurement data with • Measurement data with
location and timestamp
location and timestamp
Phase 1 • File & remote IP based
interface
Phase 2
Post Processing Tool

• connection to NMS
Data Logging Tool Field Tool Server
• Map
map data
• Network
network configuration
configuration
information
297
3G Network Optimization
3G Network optimization is divided into
1. Pre-launch optimization (without traffic, No
OSS KPI’s)
2. Continuous optimization (with traffic, KPI’s
Area
used)
3G Network optimization could be split into

Cluster
A. Cluster Optimization:
• Mainly concentrates on the detail network
optimization for each individual sub-cluster
area
• Cluster optimization work start when all the
sites in the Sub-cluster have been
implemented and integrated into the Network
• 10-12 cells form one cluster
B. Area optimization:
• takes broader approach by focusing the
network performance over the whole area
• Will begin after a number of clusters have
finished implemented & optimised 298
Pre-launch Optimization Process
Cluster Cluster Cluster Network

Preparation Tuning Optimisation Verification
Planning tool Scanner Call Call

data Measurements Performance Performance
measurements verification
for RF
optimisation for RF among clusters
verification
299
Cluster Preparation
• Cluster Identification
– Site locations, major roads, RNC borders, other geographical
aspects, e.g. rivers
– Interference analysis to minimise the amount of external
interference to a cluster from neighbouring clusters
– MapInfo plot illustrating sites within cluster, cluster borders etc.
– Site information spreadsheet (cluster id, priority etc.)
– Site Integration percentage criteria check
• Drive route planning
– A good percentage of main roads, motorways
– Different clutter types, where applicable
– Areas of special interest, e.g. airport routes, corporate routes
• Drive test equipment check
• Site Verification (crossed feeders etc.)
300
Cluster Preparation
• Collection of Cluster data
– Dominance plot
– CPICH coverage plot
– Best server Ec/Io plot
– RSCP plot
– SHO plot
– Configuration data (scrambling codes, tilts & bearing, data build)
• Network health check
– Identifying faulty Sites with
• Cell availability check
• Alarm check
– Neighbour list check (Planner vs. configuration management)
– Parameter consistency check (configuration management)
• RNW parameter check
• Transmission parameter check 301
Cluster Tuning
• RF optimisation (physical change of antenna tilt, azimuth, type and
height) based on scanner data
– Target exit criteria for RF optimisation is based on
• CPICH RSCP (Ec) per each SC
• Best Server CPICH Ec/Io
• Pilot Pollution (high RSCP vs. low EcNo)
• SHO overhead (active set count)
• Neighbour list (scrambling code) verification based on scanner data
– incl. GSM/EDGE neighbours
• Call Performance KPI Verification with logging tool
• AMR call setup success rate (CSSR)
• AMR call drop rate (DCR)
• Call setup time (CST)
302
Cluster Optimisation
• Maximise call performance KPIs by
– Analysis of Drive test data
– Analysis PM data (counters and KPIs)
• Investigation of bad quality
– lack of coverage
– external interference
– wrong or missing neighbours
• Additionally Golden cluster can be used for solving
– UE-specific problems (hanging on the cell, poor cell reselection,
poor power control)
– UE-NW incompatibilities
– Troubleshooting (call tracing)
303
Network Verification
• Used to accept the network performance based on drive tests along
”reference routes” in the area of multiple clusters
• KPIs are measured in areas of ”acceptable” coverage is achieved
304
RF & Call Performance Targets
• Minimum RF conditions for field measurements for all KPIs
(examples)
• Best server CPICH RSCP > -100 dBm
• Best server CPICH EcNo > -11 dBm
• CPICH EcNo of the 4th strongest cell > 6dB below best server
• Call Performance target examples for AMR
– Call setup success rate, 98 %
– Call drop rate, 2 %
– Overall call success rate, 96 %
– Call connection time <8 s, 98 %
305
Field Measurement Tools
• Field Measurement tools
– Scanner could be used for Coverage and Scrambling code
analysis
– Logging tools are available from a many manufacturers:
– Nemo Technologies – Layer 1 & 3 trace capability
– SwissQual - Layer 1 & 3 trace capability + subjective voice
measurement (MOS)
306
Dominance Verification
307
RSCP Verification
308
EcIo Verification
309
Pilot Pollution Verification- example
Criteria: RSCP>-92 dBm

& EcIo<-9 dB
310
Pilot Pollution Analysis I
• Pilot pollution areas can be shown with Scanner Pilot Pollution query with
user defined RSCP and EcNo levels
(CPICH_SCAN_EcNo_Sorted_By_EcNo (0)<-12)AND
(CPICH_SCAN_RSCP_Sorted_By_EcNo (0)>-92)
311
Pilot Pollution Analysis II
• Pilot polluter workbook shows the worst polluters in the area e.g. SC
344 have most of the samples
SC 344 displayed
over pilot pollution
312
Pilot Pollution Example
Scanner sees 5
SC’s, all within 5 dB
of each other.
This is clearly an
area suffering from
pilot pollution.
Looking at the dashed

lines to the cells serving at
this point we can see that
there are several cells
(SCs 136, 496, 272) which
are not particularly close
RSCP is good -91 to the point in question
dBm but EcNo is bad
-10 dB
and are candidates313 for
down-tilting
Pilot Pollution Example
• SC 272 is the furthest away these sites and is joint second best
server .
• This site is clearly
radiating much too far
and is a definite
candidate for down-
tilting.
• By removing
scrambling codes
within 5dBs of the best
server, we can
improve the best
server’s Ec/Io
SC 272
314
UL Coverage Verification
315
Throughput Verification
316
RF Optimization based on Scanner data
Coverage Data collection
A Possible Actions/Solutions
target agreed 1. Antenna Tilting
with customer 2. Antenna Panning
3. Change Antenna Type
No 4. Change Antenna Height
CPICH Ec > Threshold 5. Change CPICH Tx Pwr
e.g. –100 dBm
Currently X =
6. Add sites…
Low Max AS size = 3
CPICH Ec Yes
for Nokia RAN
A
No amount of Scrambling Yes
CPICH Ec/Io > Threshold
e.g. –11 dB Codes > X Pilot
Ec/No target agreed Pollution
with customer &
Yes
depends service mix No
No A
Dominance Area OK
Aggregated to Peak
> 3 dB Yes Multi-path
Yes Problem
A Bad No
Ec/Io
A
A
317
RF Optimization
Tuning methods for Tuning methods for
Coverage Problem Area High
HighPriority
Priority Dominance Problem Area
• Up-tilting of serving cell’s antenna to •Down-tilting of interfering cells’ antenna,

extend coverage radius and to improve which generate pilot pollution
unsatisfied coverage area •Change CPICH Tx Power: Increase
•Increase CPICH Tx Power of serving cell serving cell’s Tx power but decrease
•Change Antenna Bearing Angle: Focus interfering cell’s Tx power
the main beam of antenna to coverage •Change antenna bearing angles of cells
holes and low RSCP area involved in pilot pollution
• Change Antenna Pattern: Displace with •Change antenna patterns of cells
higher gain antenna with adequate involved in pilot pollution. Smaller gains
antenna tilting for interfering cells and higher gain for
•Increase serving cell’s antenna height to victim cell
get higher effective antenna gain but •Decrease antenna height of interfering
there is risk to make undesirable inter-cell cells and increase antenna height of
interference to adjacent cells victim cell with adequate tilting angle
318
Physical Optimisation - Antenna changes
• Antenna tilting (or panning) is needed mainly if:
– There is too much interference created by a site covering too much
(overshooting)
– There is lack of coverage or dominance
• One should carefully consider the tilt type
– Electrical or mechanical (both have advantages and disadvantages)
• Antenna tilting should be followed by another round of drive-tests in order
to evaluate the impact
• Multiple antenna tilt or azimuth changes in the same time in the same area
should be avoided
• Decisions must be made based on scanner measurements.

– UE measurements can be also used, but they can sometimes lead to wrong
results, for example in case of missing neighbours.
• Tilt changes include both down and up-tilting.
319
Electrical vs. Mechanical Tilting
Mechanical
• The downtilt angle varies over the
horizontal beamwidth. Patterns
measured ±90° from the centre of
the beam have decreasing tilt angle
until there is no tilt 90° from the main
beam.
• The horizontal half-power beam
width increases with greater downtilt
angle.
• The resulting gain reduction
depends on azimuth direction.
Electrical
• There is uniform downtilt over the
whole azimuth range.
• The horizontal half-power
beamwidth is independent of the
downtilt angle.
• There is identical gain reduction in
all azimuth directions. from Kathrein, Scala Division
320
Mechanical Tilt
• Mechanical Tilt require Site Visit
• No “real“ maximum tilt angle
• Mechanical down tilt causes deformation in the
horizontal pattern
Mechanical
Down tilt kit
Deformation of the
horizontal pattern
321
Electrical Tilt
• The Adjustable EDT antennas can be adjusted manually or remotely
• Phase shifters provides variable phase distribution which in turn keeps the
pattern shape constant
• Maximum Adjustable EDT range approx. 0-14° (normally 0-8°)
• For a higher downtilt angle a combination of the Mechanical DT and the
Adjustable EDT is recommended
Manual Remote
use use
Horizontal pattern
remains constant
322
Antenna tilt example (6 degrees downtilt)
Impact
Impacton
onEc/No
Ec/No
323
Antenna tilt example (4 degrees downtilt)
Impact
Impacton
onEc/No
Ec/No
324
Antenna tilt example -1 deg E-tilt – Overshooting Cell
No dominant
Server in the area
due the missing
site JS9218 Before After
Site JS9125 (SC28) from

cluster 5 is overshooting to Site JS9125 (SC28) was tilted325
by 1
cluster 1 deg (E-tilt From 4 deg to 5 deg)
Antenna Tilting Example - 2 deg E-tilt
• After antenna tilting, new drive-test should be performed in order to verify

changes
Before After
326
New Site Integration
• New sites should be integrated without degrading the performance of existing
sites.
• It is possible that after driving the new and surrounding site, both will require
RF optimisation (tilting, panning)
Use Both Scanner and UE

(Voice Call) Measurements to
check
• Cross Feeders
• External interference
• SHO functionality
327
New Site Integration
Different strategies can be used This option does not give us
• Integrate new sites on different frequency right information about
– Define intra-site neighbours network situation (Ec/No etc)
– Drive new site and analyse
– Apply necessary tilts to new and surrounding sites
– Add all Missing Neighbours
=> Turn Site on with correct frequency
• Integrate new site on SAME frequency during low traffic period
(Nigh Time)
This option require
• Define neighbours
possibilities to use drive test
• Turn site “on” during nigh time only team during night time (cost
• Drive new site and analyse issue)
• Apply necessary tilts to new and surrounding sites
• Add all Missing neighbours
• => Turn Site on
328
Neighbour list verification
• The neighbour list can be verified by comparing
– Best server Ec/Io results (from Scanner) with existing neighbours
e.q. with UE logging
– Best server Ec/Io results to certain Handover margin (Event 1a
criteria) based on certain number of samples
• Analysis can be done in Actix analyzer
329
Neighbour List Verification
• Neighbour definitions required by cell re-selection and handover
• Soft handovers are based upon intra-frequency neighbour list
• Hard handovers are based upon either intra-frequency (Between RNCs without Iur
or Iur congestion) , inter-frequency (IFHO) or inter-system (ISHO) neighbour lists
Intra- Inter- Inter-

Frequency Frequency System Total max 111 in RNC
Neighbors Neighbors Neighbors database, limitation
(32 /carrier)
Cell a Cell k Cell r due to specifications of
Max. 32
Max. 31
Max. 48
Cell b Cell l Cell s

Cell m Cell t SIB11/12 size
Cell c
Cell d Cell n Cell u
Cell e Cell o Cell v
• Each neighbour has a set of associated parameters e.g. CPICH measurement

offset
• The post processing tool should be able to suggest appropriate neighbour lists
• Strategy for initial system deployment is to place the emphasis upon adding
neighbours rather than removing them
330
Max. 96
Neighbour List Verification – SIB 11/12 neighbours due
SIB limitation
There is a restriction on the number of cells contained in SIB 11/12

due to an inconsistency problem in the standards
• SIB 11/12 should be able to contain a maximum of 96 neighbours
– (32 intra-frequency cells, 32 inter-frequency cells and 32 GSM cells)
• On the other hand, the physical size of SIB data (no more than 3552 bits)
has capacity only for about (depending on the type etc.) 47 cells!!
• If too many adjacencies are declared, the cell will go blocked by system with
alarm:
– 7771 WCDMA CELL OUT OF USE (BCCH scheduling error)
• As a rule of thumb, assuming that …
ADJS=15, ADJG=15, ADJI=15
… and “realistic worst case values”, SIB11 length = 3187.5 < 3552 -> OK!!
• Some sites might need additional neighbors and might pose a problem with the
SIB11 limitation
– Avoid setting AdjsQoffset2 values, different CPICH values or other parameters
used to tune cell reselection or handover
Further information Technical Note No. 046 / Restriction on number of
cells in SIB11/12 due to inconsistency problem in 3GPP TS 25.331 331
Neighbour List Verification – SIB 11/12
• When offsets are added to the neighbours or the CPICH power of the
neighbour cell is different to that of the serving cell the length of the
message for each neighbour in SIB 11/12 is increased
– The length of one ADJS with no offsets is 48 bits
– The length of an ADJS with AdjsQoffset1 or AdjsQoffset2 is 48 or 56 bits
on a case by case basis (average length 55.2 bits)
– The length of an ADJS with AdjsQoffset1 and AdjsQoffset2 is 56 or 64
bits on a case by case basis (average length 62.1 bits)
• When the neighbour cell has a different CPICH Tx Power from the
serving cell
– Without Offset and AdjsPtxCPICH : 54.2bits
– With AdjsQOffset1 or AdjsQOffset2 : 61.1bits
– With both AdjsQOffset1 and AdjsQOffset2 : 68bits
332
• There are a number of approaches that can be used to both plan and
verify the neighbour plan
Neighbour Creation Neighbour

Verification
Neighbour
Manual Check Analytical
Creation
Planning Tool
Other
Manual Check
Drive Testing
Measured
Network Stats
333
Neighbour List Verification - Analytical
• Analytical Neighbour Planning methods are traditionally used to

generate original neighbour lists but they can also be used for
verification and new site addition.
• Planning Tools – use coverage arrays and handover parameter data
to determine required neighbour relations
• Other desk top tools can be based on:
– Site coordinate data,
– Cell azimuth data,
– Cell coverage distance estimations
– Antenna beam widths
334
Neighbour List Verification - Manual
• Cell in blue had nine 3G

neighbours defined –
highlighted in green on
map
• 6 missing neighbours
were identified for this
site and added – Manual
Check would have
identified at least 5
335
Neighbour List Verification – Drive Test
• Process starts directly with the drive test
High Level
Process
Start Record drive test
results with Scanner in
‘TOP N’ mode
Input Data Source

CPICH Scrambling code Ec/Io Scanner
Measurement position Scanner
Cell ID, cell position, cell azimuth Planning Tool
Cell scrambling code Radio Design
Cell neighbour list Planning Tool
336
• Part automatic and part manual neighbour list analysis
• Analysis within a tool such as Actix Analyzer is automatic and numerical
• Analysis requires the definition of a neighbour window which is applied to the
CPICH Ec/Io measurements
• Recommended to set the neighbour window as 10 dB (drop window + margin)
Strongest Neighbours
Ec/Io High Level Process
Ec/Io reported when
within this range Record drive test
Start
results with Scanner in
‘TOP N’ mode
Neighbou
r Window
Analysis (within
Actix Analyzer) Finish
Time
SC Site Sector Num Of Samples Suggested NBR Additions
216 Ewer Park Stud 00161126C 97 85 (47.4%); 303 (17.5%)
349 Reading Rubgy FC 00179719B2 97 351 (28.9%)
83 WOODLEY ATE 00001085A 96 350 (34.4%); 84 (25.0%)
436 TILEHURST ATE 00001081B 90 434 (22.2%); 283 (22.2%)
337
Numerical Analysis
• Analysis provided by the post

processing tool: Start
Run
Run (Actix)
(Actix)
• Consider adding neighbours Analyzer
Analyzer Routine
Routine
reported by the tool

Consider
Consider adding
adding Is
Is the
the
No
suggested
suggested Neighbour
Neighbour List
List
neighbours
neighbours full?
full?
• If neighbour list is full then
consider replacing some of Consider
Consider
the existing neighbours replacing
replacing
existing
existing
neighbours
neighbours by by
those
those suggested
suggested
• Do not remove existing by
by the
the tool
tool
neighbours without further
investigation
Progress to
further additions
338
Example for Reading Central (SC 280)
% Scrambling Code Samples
• Analysis output 40.00

generated by Actix 35.00
Analyzer
30.00
25.00
% of Samples
• Bar chart 20.00
33.90
• List 15.00
18.50
10.00
13.80
9.40
5.00
3.90
3.10
2.00
1.20
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
422 423 420 283 282 286 348 177
Neighbour Scram bling Code
SC Site Sector Num Of Samples Suggested NBR Additions

216 Ewer Park Stud 00161126C 97 85 (47.4%); 303 (17.5%)
349 Reading Rubgy FC 00179719B2 97 351 (28.9%)
83 WOODLEY ATE 00001085A 96 350 (34.4%); 84 (25.0%)
436 TILEHURST ATE 00001081B 90 434 (22.2%); 283 (22.2%)
Current NBR List (From CellRef File) NBR Not Seen

{}
{} 404 (0.1%); 287 (0.1%); 286 (0.1%)
{} 457 (0.1%); 456 (0.1%); 405 (0.1%)
{}
339
• Subsequent manual
tuning of neighbour list Continue from Analysis (slide 49)
• As a minimum define the

neighbour list as Complete
Complete visual
visual
Add
Add inspection
neighbours inspection of
of
All cells belonging to that site neighbours neighbour
neighbour list
list
The first perimeter of cells
Does
Does list
list include
include all
all
NO
cells belonging
cells belonging to to
Cell for which that
that site
site and
and the
the
the neighbour first
first perimeter of
perimeter of
list is being cells
cells
defined
Yes
Add
Add any
any other
other
neighbours
neighbours which
which
Neighbours may
may improve
improve
coverage
coverage
Finish
340
Neighbour List Verification – Co Scrambling Code
Problem
• An important part of neighbour list verification (against the assigned
scrambling codes) is to ensure that the neighbour lists do not pose
any scrambling code clashes.
• A possible impact is that the network attempts to set up a link on a
different cell (scrambling code) to that measured by the UE, or due to
combining of the neighbour lists by the RNC duplicate scrambling
codes may be introduced, and the call may drop.
341
• Scenario 1 Problem
– In some cases, the neighbour definition is created to a cell (Cell A), which is far from
the source cell, and having the same scrambling code as Cell C, which is very close
and is not defined as a neighbour.
– This problem is caused by errors in mapping the Primary Scrambling Code Picked by
the scanner to a certain cell ID. The target Primary Scrambling code (123 in the
example) was picked by the scanner, but was associated with a wrong CID, Cell A
instead of Cell C.
– Prevention of the problem is done through the following steps:

• Neighbour relationships cannot be defined without visual check on the map
• Long distance neighbours (over ? km) are audited by querying the RNW
configuration database and corrected after visual checking on map
• Situations where the geographical distance to a neighbour is more than the
geographical distance to another cell with the same scrambling code (which is not
a defined neighbour) are also audited by querying the RNW configuration
database and corrected
342
Problem
• Scenario 2 (two-way SHO)
– In this scenario the UE is connected to two cells (Cell B and Cell C in the example).
These two cells have ADJS relationships defined with a different Cell ID but with the
same scrambling code. While the UE is in soft handover, the RNC combines the
neighbour list of Cell B and Cell C.

• All mechanisms described for Scenario 1 will help to eliminate the long distance
adjacency between Cell A and Cell B
• The scenario shown above can be audited by querying the RNW configuration
database and corrected after visual checking on map
343
Problem
• Scenario 3 (three-way SHO)
– In this scenario Cell B and Cell D are not defined as neighbours. However, they are
both in the active set due to ‘three-way’ soft handover between Cell B, C and D, and
thus the combination of the neighbour lists will happen in the RNC.

• All mechanisms described for Scenario 1 will help to eliminate the long distance
adjacency between Cell A and Cell B
• The scenario shown above can be audited by querying the RNW configuration
database and corrected after visual checking on map
344
Part XII
Radio Resource Management
345
346
347
348
349
RRM Control Processes
350
WCDMA Radio Resource Management:
Logical Model
• RRM is made up of a number of closely interdependent functions (i.e.
algorithms)
• These functions can be divided into;
LC
• Cell Based PS
– Load control (LC)
RM
– Admission control (AC) AC
– Packet scheduling (PS)
– High Speed Downlink Packet Access Cell based functions
– Resource manager (RM)
PC
• Connection Based
– Handover control (HC) HC
– Power control (PC)
Connection based functions

351
RRM control processes
• Admission control
– Performs the admission control for new bearers to enter/leave
the network.
– Predicts the interference caused by the bearer and checks
whether there is room for it.
– Power allocation
• Packet Scheduler
– Scheduling packets to the radio interface (UL/DL)
• Load Control:
– Takes care of radio network stability
– Gathers interference information and produces a load vector
• Resource manager
– Manages the physical resources of RAN and maintains the
code allocation
352
RRM control processes
• Power Control
– Closed loop PC compares the measured SIR with SIR-target
and accordingly transmits an up/down PC command at 0.667
ms interval
– Open loop PC estimates the needed power based on pathloss
+ interference measurements (RACH).
– Outer loop PC sets the SIR target for the fast closed loop PC
• Handover Control
– Soft (intra-frequency) handovers: softer between cells within
one BS, intra-RNC soft, inter-RNC soft
– Inter-frequency (hard) handovers: Intra-BS, Intra-RNC, Inter-
RNC (-MSC)
– Inter-RAT handovers: WCDMA <-> GSM
353
Power Control
Power Control loops in WCDMA
MS BTS RNC
Open Loop Power Control (Initial Access)
Closed Loop Power Control
354
Power Control Loops
• Effective power control is essential in WCDMA due to frequency re-
use factor of one (in ideal case)
• Closed loop e.q. Fast power control
– Makes Eb/No requirements lower
– Equalizes received powers at BTS in uplink (avoids near-far effect)
– Introduces interference peaks in the transmission
• Open loop power control for initial power setting of the UE
• Outer PC loop at a slower rate, across the Iub interface in uplink
– At a much slower rate, across the Iub interface in uplink
– Adjusts the SIR target to achieve a target BLER
– Also similar outer loop power control in MS
– There is also similar outer loop power control in UE
355
Power Control & Diversity
• At low UE speed, power control compensates the fading : fairly

constant receive power and Tx power with high variations
• With diversity the variations in Tx power is less
• At UE speed >100kmph fast power control cannot follow the fast
fading, therefore diversity helps keep receive power level more or
less constant
• In the UL Tx affects adjacent cell interference and Rx power affects
interference within the cell.
356
Admission Control & Packet Scheduler
• AC handles new incoming traffic to the RAN by

– estimating the total load caused by adding a new RAB in uplink
and downlink
– and decides whether or not this can be admitted.
• AC also sets :
– initial DL transmission power for the channel
– the power control range as well as many other parameters
(UL/DL BLER, Eb/No Target, SIR Target, e.g. Transport Format
Set.
• PS handles all the NRT data connections. PS is determining the
time a packet is sent and which bit rate is used.
357
Admission Control & Packet Scheduler
• The key function of AC and PS is to maximize capacity

(throughput) by estimating the load and to fill the system up to
maximum loading while still ensuring the required quality of service
for RT traffic.
• In uplink, the basic measured quantity indicating load is the total
received power of a BS, PrxTotal
• In downlink, the basic measured quantity indicating load is the total
transmitted power of a BS, PtxTotal
358
Admission Control
Uplink admission control
• In uplink the total received wideband interference power measured
indicates the traffic load of the radio resources .
• The fundamental criteria of evaluation is based on
I total_old + ∆ I < I threshold

• Ithreshold indicates the traffic load of the radio resources
• In uplink, the total received power is the function of the maximum
interference received in the wideband spectrum .
power
max planned power Ithreshol
d
∆I =? max planned load

Itotal_old
359
load
Admission Control
Uplink admission control
UL interference power
Prx_target_BS
Marginal load area Prx_offset
Prx_target
TRHO_threshold
planned uplink
interference power
Planned load area
Load
Prx_target defines the optimal operating point of the cell interference power, up to
which the AC of the RNC can operate.
360
Admission Control
Downlink admission control
DL transmission power
Ptx_target_BS
Marginal load area Ptx_offset
Ptx_target
TRHO_threshold
Planned load area

planned Downlink
interference: carrier
transmission power
Load
Downlink power increase estimation is done for non-controllable load just like UL
power increase.
361
Admission Control Functional Overview
• Maximises capacity whilst maintaining stability
• Decides if new RAB request is admitted into RAN

RNC
• AC tasks LC
– A) Admission decision PS
• RRC connection
RM
AC
• (RAB) RT connection
• (RAB) NRT connection Network based functions
– B) Definitions of RB configuration
– C) Outer loop power control configuration
– D) RL power allocation
• initial, minimum and maximum transmission power
362
Admission Control Functional Overview
RRC Connection
Establishment
0
NRT Admission Decision

Process (PS)
6 RAB
Establishment
Core
UL/DL Load Change Nokia RAB ASSIGNMENT REQUEST
Report to LC NRT PS RNC Network
5 call
RAB AC
admitted
RT CS call Radio Access Bearer
Admission Service Request
Decision Quality
1
4 Requirements of
Power Increase
Estimates 2 Radio Bearer RAB attributes (HLR);
3 • SDU error ratio
RAB request • traffic class
denied • max bit rate
Outer
Outer loop
loop PC
PC RB
RB configuration
configuration
configuration
configuration (RNC);
(RNC); (RNC);
(RNC);
•• target
target BLER
BLER •• RLC
RLC mode
mode
•• target
target Eb/No
Eb/No •• TrCH
TrCH parameters
parameters
•• initial
initial SIRtarget
SIRtarget
363
Admission control for NRT RAB
• AC allocates a zero bit rate for the NRT RAB

– PS allocates the bit rate to the radio bearer after reception of capacity
request
PS checks load and makes estimate of the load change
• No power based decision is made in AC decision for NRT RAB
• AC produces the TFS for NRT RBs based on RAB requirements

– AC will check that the maximum bit rates requested in the RAB Assignment
Request message are lower or equal than operator configurable limits for
the specific target cell
• QoS negotiation RAB bit rate is decreased to the cell maximum bit
rate
• MaxBitRateDLPSNRT and MaxBitRateULPSNRT define the
maximum allowed bit rates in a cell
– range: [8, 16, 32, 64, 128, 256, 320, 384] kbps, default: 384 kbps
364
UL Cell Load Admission Criteria Thresholds
• The estimated non-controllable UL interference power is compared

against the threshold
Prx_nc + ∆Prx_nc < Prx_target Prx_target =

PrxNoise + PrxTarget
If the equation is NOT satisfied then the RAB can NOT be admitted!
• If there is queuing time left for the RAB (parameter AcRabQueueTime) then
the RAB remains in the queue to wait for the next PrxTotal measurement,
otherwise it is removed from the queue
365
Parameters
• When the non-controllable part of the cell load (Prx_NC) exceeds the target
limit, AC will reject those RAB establishment requests that would mean
immediate UL load increase (i.e. RT services)
• The packet scheduler (PS) is responsible on handling the NRT bearer
– When load is too high bit rates are decreased etc.
• When the PrxTarget level

has been exceeded
PrxNoise [dBm] + PrxTarget [dB] + PrxOffset [dB] – it is possible for the
PrxTotal AC/PS to admit NRT
[dBm] services (AC)
Prx_NC – it is possible to schedule
[dBm] (allocate) bitrate for NRT
PrxNoise [dBm] + PrxTarget [dB] services
– bitrate (load) increase
for NRT is not possible
Note:
• PrxTarget_BS is defined as PrxTarget + PrxOffset
• PtxTarget_BS is defined as PtxTarget + Ptx Offset
366
UL Admission Decisions
Prx_Target_BS
Prx_Target
Controllable load
Controllable load
∆Prx_NC ∆Prx_NC
Prx_NC Prx_NC
Case I Case II
Prx_NC ++ ∆Prx_NC
Prx_NC ∆Prx_NC ≤≤ Prx_Target
Prx_Target Prx_NC ++ ∆Prx_NC
Prx_Target
RT RAB admitted (as well as NRT) RT RAB admitted (as well as NRT)
Priority based scheduling by PS 367
UL Admission Decisions
Prx_Target_BS
Controllable load
∆Prx_NC
Prx_Target
Controllable
load
∆Prx_NC
Prx_NC
Prx_NC
Case III Case IV
Prx_NC ++ ∆Prx_NC
Prx_Target Prx_NC ++ ∆Prx_NC
Prx_NC ∆Prx_NC >> Prx_Target
Prx_Target
RT RAB denied (NRT permitted) RT RAB denied (NRT RAB permitted)

Overload control by PS RT over NRT pre-emption 368
DL Cell Load Admission Criteria Thresholds
• For the RB to be established the increase of the non-controllable load
∆Ptx_nc is estimated on the basis of the measured P tx _ total
• Then the estimated non-controllable DL power is compared against the
threshold
• If the equation is NOT satisfied then the RAB can NOT be admitted!
Ptx_nc+ ∆Ptx_nc< Ptx_targe
Ptx_nc = Ptx_total – Ptx_nrt ∆Ptx _ nc = ∆Ptx + ∆Ptx,inactive
369
DL Admission Decision
Ptx_Target_BS
Ptx_Target
Controllable load
Controllable load
∆Ptx_NC ∆Ptx_NC
Ptx_NC Ptx_NC
Case I Case II
Ptx_nc + ∆Ptx_nc < Ptx_target Ptx_nc + ∆Ptx_nc < Ptx_target
RT RAB admitted (as well as NRT) RT RAB admitted (as well as NRT)
370
DL Admission Decision
Ptx_Target_BS Controllable load

Controllable load ∆Ptx_NC
Ptx_Target
∆Ptx_NC
Ptx_NC
Ptx_NC
Case III Case IV
Ptx_nc + ∆Ptx_nc < Ptx_target Ptx_nc + ∆Ptx_nc < Ptx_target
RT RAB admitted (as well as NRT) RT RAB denied (NRT RAB permitted)
Overload actions started by PS RT over NRT pre-emption 371
Packet Scheduler
• Packet scheduler is a general feature, which takes care of

scheduling radio resources for NRT radio access bearers for both
uplink and downlink.
• The packet access procedure in WCDMA should keep the
interference caused to other users as small as possible.
• Packet access is implemented for both dedicated (DCH) and
common control transport channels (RACH/FACH).
• There are three scenarios for WCDMA packet access:
• infrequent transmission of short packets,
• frequent transmission of short packets (RACH/FACH)
• transmission of long packets (DCH)
• Packet scheduler makes the decision of the used channel type for
downlink direction. For uplink direction the decision of the used
channel type is made by UE
372
Why Packet Scheduling?
• Suitable for controllable load (=best effort NRT)
• Requires fast resource allocation
Planned
Load Load/Power
(~ Power) Free Capacity Target
HSDPA
reservation
(when HSDPA
preferred over
NRT DCH)
Non-Controllable Load
• real time traffic
• interference from other cell users
• noise
time 373
Why Packet Scheduling?
It is characteristic for RT traffic that it’s load cannot be controlled in efficient way. Load caused
by RT traffic, interference from other cell users and noise together is called non-
controllable load. The available capacity, which is not used by non-controllable load, can
be used for NRT radio bearers on best effort basis. To fill the whole load budget and
achieve the maximum capacity, the allocation of NRT traffic needs to be fast.
The Packet scheduler is a general feature, which takes care of scheduling radio resources for
NRT radio bearers for both uplink and downlink directions; Packet scheduling happens
periodically (with the period of tens of milliseconds) and is implemented for both dedicated
(DCH) and common control transport channels (RACH/FACH).
Scheduled capacity depends on the UE capabilities, Node B capabilities, current load of the
cell as well as the availability of the physical radio resources.
Packet scheduler and MAC layer together make the decision of the used channel type for
downlink direction, data transmission on dedicated channel is initiated when MAC layer
requests transmission capacity
For uplink direction the decision of the used channel type is based on UE measurements
and parameters controlled by network. Data transmission on dedicated channels is initiated
when a capacity request is received from UE.
The selection of the channel type is done fast - taking into account the data amount in the
buffers and the current radio conditions
374
Packet Scheduler
• The packet scheduler takes care of scheduling radio resources for non-real
time radio bearers for both the uplink and the downlink directions
• Packet access is implemented for dedicated (DCH) as well as common

control transport channels (RACH/FACH)
– Packet access is implemented for high speed downlink shared channel
(HS-DSCH) in the case of HSDPA
• The packet scheduler and the medium access control (MAC) layer together
make the decision regarding which type of channel to use in the downlink
direction
375
Packet Scheduler as part of RRM
• The packet scheduler (PS) co-operates with other radio resource management
functions like handover control (HC), load control (LC), admission control (AC) and
the resource manager (RM)
– HC provides active set information
– LC provides periodical load information to PS on a cell basis
– PS informs AC and LC of the load caused by non-real time radio bearers
– AC informs PS when new non-real time radio bearers are admitted,
reconfigured or released
– RM allocates the RNC internal resources, downlink spreading codes and takes
care of allocating radio links using the base station application protocol (NBAP)
– RM also takes care of transport resource reservation for non-real time radio
bearers using transport resource manager (TRM) services. RM actions are
done when requested by the PS
– The radio resource control (RRC) protocol takes care of L3 signalling between
the RNC and the UE
• The L3 RRC signalling needed by the PS includes uplink capacity
requests and channel allocation procedures in both directions (uplink and
downlink)
– The medium access control (MAC) protocol produces radio bearer-specific
downlink capacity requests to the PS according to the radio link control (RLC)
buffer levels in the RNC
– MAC also sends activity and inactivity indications on an RB basis. 376
Packet Scheduler functions
• Packet Scheduler consists of multiple different functions which can be

categorised based on the scope of the function
– UE-specific part
• Functions working based on single radio link/bearer status,
measurements and conditions
– Cell-specific part
• Functions working based on cell level measurements and conditions
UE Specific: Cell Specific:

• Capacity Request & Traffic Volume • Cell Load measurement and
Measurements averaging
• NRT DCH throughput • Queuing of capacity requests
measurements and minimum • Channel type selection
utilisation
• Power budget for packet scheduling
• Flexible upgrade of the NRT DCH
and High throughput measurement • Bit rate allocation process/method
• Load increase - Bit rate upgrade
• Dynamic link optimisation for non-
• Load decrease - Overload control
real time traffic coverage
• Enhanced priority based scheduling
• PS interruption timer
377
Supported Radio Bearers in RAN
Future
RAN04 RAS05 RAS05.1
items
AMR
AMR 12.2
12.2
Transparent
Transparent CS
CS data
data 64
64
Non-transparent
Non-transparent CSCS data
data 14.4,
14.4, 57.6
57.6
NRT
NRT PS
PS data
data 8,
8, 16,
16, 32,
32, 64,
64, 128,
128, 384
384 (DL)
(DL) (CELL_DCH)
(CELL_DCH)
NRT
NRT PS
PS data
data 384
384 (UL)
(UL) (CELL_DCH)
(CELL_DCH)
NRT
NRT PS
PS data
data 256
256
(CELL_DCH)
(CELL_DCH)
NRT
NRT PS
PS data
data 16
16 (UL),
(UL), 32
32 (DL)
(DL) (CELL_FACH)
(CELL_FACH)
Streaming
Streaming PS
PS data
data 8,
8, 16,
16, 32,
32, 64,
64, 128,
128, 256
256 (DL)
(DL)
NRT
NRT PS
PS data
data on
on HS-DSCH
HS-DSCH 1800
1800
RAS05 CD HS-DSCH
HS-DSCH 3600
3600
AMR = Conversational CS speech Lower

Lower AMR
AMR speech
speech rates
rates
Transparent CS data = Conversational CS
data
WB-AMR
WB-AMR speech
speech rates
rates
Non-transparent CS data = Streaming CS
data Conversational
NRT PS data = Interactive/Background PS Conversational
PS
PS QoS
QoS
data 378
Supported Radio Bearers in RAN for Multicall
RAN04 RAS05 RAS05.1
2-3
2-3 NRT
NRT PS
PS data
data
MultiRAB: AMR + 1-3 PS data AMR
AMR 12.2
12.2 ++ 1-3
1-3 NRT
NRT PS
PS data
data
MultiRAB: CS data + PS data Transparent
Transparent CS
CS data
data 64
64 ++ NRT
NRT PS data 88 1)
PS data 1)
Non-transparent
Non-transparent CS
CS data
data 57.6
57.6 ++ NRT
NRT PS
PS data
data
Non-transparent
Non-transparent CS
CS data
data 14.4
14.4 ++ NRT
NRT PS
PS data
data
Streaming
Streaming PS PS data
data ++ NRT
NRT PS data 88 1)
PS data 1)
AMR
AMR 12.2
12.2 ++ streaming
streaming PSPS data
data ++ NRT
NRT PS
PS
data 88 1)
data 1)
AMR
AMR ++ NRT
NRT PS
PS
data
data on
on HS-DSCH
HS-DSCH
1) The total aggregate user bit rate of the RRC connection must not exceed 384
kbit/s (AMR+PS384 not possible)
(The restriction does not apply to HSDPA.) 379
Bit rate upgrading
• The dedicated channel of a non-real time (NRT) radio access bearer (RAB)
can be upgraded due to
1) High amount of data in buffer – Capacity request/ Bit rate upgrade
2) High utilisation/throughput – Flexible upgrade of the NRT DCH data rate
• It is possible to upgrade the NRT DCH data rate from any bit rate below the
maximum allowed bit rate to the maximum allowed bit rate (Flexible upgrade
feature)
• The dedicated channel upgrade procedure is performed in CELL_DCH state

and it requires the reconfiguration of radio link, transmission and RNC
internal resources
380
Bit rate downgrading
• The dedicated channel of a non-real time (NRT) radio access bearer (RAB)
can be downgraded or released due to multiple causes
1) Excessive downlink power – Dynamic link optimisation (Dylo) for
non-real time traffic coverage feature
2) Different congestion situations – Enhanced priority-based scheduling
(EPBS) and overload control
3) Low utilisation/throughput – Throughput-based optimisation of the
packet scheduler
4) Maximum bit rate limitation – Another RAB is setup for the same UE
5) Inactivity of the radio bearer
• The dedicated channel downgrade procedure is performed in CELL_DCH

state and it can be performed by
– Reconfiguration of radio bearer (also transmission, WBTS and RNC
resources) Limitation of the maximum transport format combinations
(temporary); TFC
381
Packet Scheduler actions during call – Unloaded cell
AC - AC makes admission decision and allocates a 0 bit rate to the NRT radio bearer
PS1 - “Bit rate allocation”, after receiving a capacity request PS allocates initial bit rate
PS2 - “Flexible upgrade”, After receiving a capacity request PS allocates maximum (high) bit
rate
PS3 - “Throughput-based optimisation“, PS performs radio bearer reconfiguration to lower bit
rate
PS4 - “Flexible upgrade”, PS performs radio bearer reconfiguration to higher bit rate
PS5 - “RRC state transition”, PS initiates state transition to CELL_FACH due to inactivity
PS2 PS3 PS4 PS5
Max. bit rate
Allocated bit rate
PS1
Initial bit rate
Actual throughput
AC
0 bit rate
382
Packet Scheduler actions during call – Loaded cell
AC - AC makes admission decision and allocates a 0 bit rate to the NRT radio bearer
PS1 - “Priority based scheduling”, after receiving a capacity request PS allocates initial bit rate
after downgrading an existing bearer (load margin)
PS2 - “Flexible upgrade”, After receiving a capacity request PS allocates higher bit rate (normal
load)
PS3 - “Enhanced Overload control“, PS performs radio bearer reconfiguration to until

minimum bit rate (overload)
PS4 - “Flexible upgrade”, PS performs radio bearer reconfiguration to higher bit rate (normal
load)
PS5 - “RRC state transition”, PS initiates state transition to CELL_FACH due to inactivity
383
Packet Scheduler actions during call – Loaded cell
Overload
Load Margin
Normal load
PS4 PS5
Allocated bit rate
Max. bit rate
PS2 PS3
PS1
Initial bit rate
AC
Minimum bit rate
Actual throughput
PBS FLXU EOLC FLXU 384

RRC State Machine
Frequent cell updates
Cell Update, UL Tx
UL Tx
UL/DL activation timer
UTRA RRC Connected Mode
UE in DRX mode UE in DRX mode
discontinous reception
URA_PCH CELL_PCH discontinous receiption
Dedicated resources Common resources

allocated (DCH) CELL_DCH CELL_FACH allocated (RACH-FACH)
Tx and Rx mode Tx and Rx mode
Inactivity Timer
Overload Traffic volume
RACH load
Cell selection
Cell re-selection Idle Mode
385
Packet Scheduler
Load Decrease Example
386
Packet Scheduling Principle
387
Load Control
Capacity
• The traffic can be divided into two groups
– Real Time (RT)
– Non-Real Time (NRT)
• THUS some portion of capacity must be reserved for the RT traffic
for mobility purposes all the time. The proportion between RT and
NRT traffic varies all the time.
Overload area
Overload Margin
Load Target Estimated capacity for
NRT traffic.
Power
Measured load caused

by noncontrollable load
Time 388
Load Control
Definition of Non-controllable traffic
• Since it is not enough to divide the load to RT and NRT one must take
into account the interference coming from surrounding cells.
Traffic is divided into controllable and non-controllable traffic.
Non-controllable traffic = RT users + other-cell users +

noise + other NRT users which
operate at minimum bit rate
Controllable traffic = NRT users
389
Logical description of load control
• The purpose of load control is to optimize the capacity of a cell and

prevent overload situation.
• Load control consists of Admission Control (AC) and Packet
Scheduler (PS) algorithms, and Load Control (LC) which updates the
load status of the cell based on resource measurements and
estimations provided by AC and PS.
Load change
info
AC
Load status
PS
LC NRT load
390
Handover Control
391
Handover Control (HC)
• HC is responsible for
• Managing the mobility aspects of an RRC connection as UE move around network
• Maintaining high capacity by ensuring UE is always served by strongest cell
Soft(er) Intra-layer
Intra-frequency Inter-layer
Handover
Intra-System
WCDMA to WCDMA
Inter-frequency Intra-layer
Hard Handover Inter-layer
Intra-frequency Intra-layer
Inter-layer
Requires
Compressed Mode
or Dual Receiver
Inter-System UE
(Inter-RAT)
WCDMA to GSM
WCDMA to GPRS
GSM to WCDMA Inter-frequency Intra-layer
GPRS to WCDMA Hard Handover Inter-layer
392
Softer Handover
• Handover between cells
within a BS
• softer handover is handled
by BS internally
Sector/Antenna
RAKE • softer handover probability
combining about 5 - 15 %
(MRC) • no extra transmissions
across Iub
• basically same RAKE MRC
processing as for
multipath/antenna diversity
(BS / MS). More RAKE
fingers needed.
• provides additional diversity
gain
• softer handover does create
RNC additional interference and
needs BS PA resources
393
Soft handover
• Handover between cells from different

BS's
• Soft handover probability about 20 - 50
%
• Required to avoid near/far effects
• Extra transmission across Iub, more Except for the TPC symbol
exactly the same information
channel cards are needed (symbols) sent over air.
fo
RNC
r eli ability in Differential delay in order
CN frame of fraction of symbol duration
frame
selection /
duplication
• DL/MS: Maximal ratio combining fra

m
• UL/RNC: Frame selection combining e
re
lia
• Soft handover does create additional bi
lit
interference in downlink and needs BS yi
nf
power amplifier resources o
• DL Power drifting in soft HO BSs a

problem due to independent errors in
uplink commands
394
Handover Types
• Intra-Frequency Handovers
– Softer Handover
• Handover between sectors of the same Node B (handled by BTS)
• No extra transmissions across Iub interface
• Maximum Ratio Combining (MRC) is occurring in both the UL and DL
– Soft Handover
• MS simultaneously connected to multiple cells (from different Node Bs)
• Extra transmission across Iub, more channel cards are needed
(compared to non-SHO)
• Mobile Evaluated Handover (MEHO)
• DL/UE: MRC & UL/RNC: Frame selection combining
– Hard Handover
• Arises when inter-RNC SHO is not possible (Iur not supported or Iur
congestion)
• Decision procedure is the same as SHO (MEHO and RNC controlled)
• Causes temporary disconnection of the (RT) user
• Inter-Frequency Handover
– Can be intra-BS, intra-RNC, inter-RNC
– Network Evaluated Handover (NEHO)
– Decision algorithm located in RNC
• Inter-RAT Handover
– Handovers between GSM and WCDMA (NEHO) 395
Soft Handover
WCDMA F1 WCDMA F1 WCDMA F1
• Main method is soft/softer Handover

• MS connected to two or more cells
• Up to 3 branches can be used (MaxActiveSetSize parameter)
• Seamless Handover without any disconnection of UE-RAN connection
• softer Handover probability about 5-15 % : soft Handover probability about 20-50 %
• Same power levels in use for all soft Handover branches
• Optimum power levels for link quality and cell capacity
• Macro Diversity Combining (MDC) gain to reduce power levels
• Mobile Evaluated Handover (MEHO)

• Enables different cell sizes due to Common Pilot Channel Ec/No measurement
• Optimum fast close loop power control in use
• Outer loop PC after MDC controls the BLER according to set target and thus maintains
the quality
396
Soft Handover
• HC supports the following measurement reporting events and features:

– Event 1A: A primary CPICH enters the reporting range (Ncell addition)
– Event 1B: A primary CPICH leaves the reporting range (Ncell deletion)
– Event 1C: A non-active CPICH becomes better than an active primary
CPICH (Ncell replacement)
– Cell individual offsets for modifying measurement reporting behaviour
– Mechanism for forbidding a neighbouring cell to affect the reporting
range
• Handover decision performed by RNC based on measurements and available

resources
• Admission Control can reject the branch addition in case the maximum load
is achieved in DL (threshold + offset), valid both for RT and NRT bitrates.
• Hard blocking may prevent branch addition
397
1A: A Primary CPICH Enters the Reporting Range
1 • Reporting range defined by offset (point 1 on next slide) from either
(i) best CPICH Ec/Io in the AS, or
(ii) sum of the AS measurement results
set using ActiveSetWeightingCoefficient (FMCS) (0 = use best CPICH : <>0 = use
sum)
• Offset defined using AdditionWindow (FMCS)
2 • Event 1A triggered when CPCIH3 Ec/Io enters UE reporting range (point 2)
3 • If CPICH3 Ec/Io remains within reporting range for a time period defined by
AdditionTime (FMCS) , and the AS is not full, UE sends measurement report to RNC
which adds Ncell3 to AS if possible (point 3)
• AdditionTime defines the 'time-to-trigger' interval between the Ncell first entering the
reporting range and the UE sending the measurement report to the RNC
4 • If RNC is unable to add Ncell to AS, UE will wait for a period of time, defined by
AdditionReportingInterval (FMCS) (point 4) after the first measurement report,
before sending further reports periodically, with interval AdditionReportingInterval,
until (a) Ncell moves out of reporting range, or (b) RNC adds Ncell to AS.
398
1A: A Primary CPICH Enters the Reporting Range
Ec/Io ActiveSetWeightingCoefficient
Strongest CPICH in AS P CPICH 1
P CPICH 2 AdditionWindow
1
P CPICH 3
time
3 AdditionTime
Measurement AdditionReportingInterval 4
no
Report
Add to
the AS?
399
RNC
1B: A Primary CPICH Leaves the Reporting Range
1 • Event 1B triggered when CPICH3 EC/I0 drops out of the UE reporting range for a
defined period of time (point 1 on next slide)
2 • Reporting range = Strongest CPICH in AS - DropWindow (FMCS) (point 2)
3 • Time-to-trigger period set using DropTime (FMCS) (point 3)
• UE sends RNC measurement report on expiry of DropTime to remove Ncell

(CPICH3) from AS.
• RNC drops the cell from the AS and UE drops the cell from the AS to the
Neighbour Set
400
1B: A Primary CPICH leaves the Reporting Range
Ec/Io Strongest CPICH in AS

P CPICH 1
DropWindow
P CPICH 2 2
P CPICH 3
time
DropTime
Remove the
Measurement reported cell
Report from the AS
401
1C: A non-active CPICH becomes better than an
active primary CPICH
• UE AS full (MaxActiveSetSize = 3)
1 • Event 1C triggered when CPICH4 EC/IO > CPICH3 in AS by a defined margin,
ReplacementWindow (FMCS). (point 1 in next slide)
• Sending of measurement report to RNC can be delayed by using 'time-to-trigger'
period set by ReplacementTime (FMCS)
2 • If 'time-to-trigger' is used difference between CPICH4 and CPICH3 must be ≥
ReplacementWindow for the period of ReplacementTime (point 2)
3 • If the RNC is not able to replace the Ncells, the UE continues to send
measurement reports periodically, with interval ReplacementReportingInterval,
to the RNC until (a) CPICH4 falls out of ReplacementWindow or (b) RNC
replaces CPICH4 by CPICH3 (point 3)
402
1C: A non-active CPICH becomes better than an
active primary CPICH
Ec/Io AS has 3 cells
P CPICH 1
P CPICH 2
P CPICH 4
1
ReplacementWindow
P CPICH 3
weakest CPICH3 in AS
time
2 3
ReplacementTime ReplacementReportingInterval
no
Measurement AS
Report update?
403
Company Confidential RNC
Individual Ncell Offset

• cell individual offsets for modifying measurement reporting behaviour
• effectively 'moves' cell border (shrinks or enlarges cell)
• offset set using AdjsEcNoOffset (= CPICH Ec/No Offset) (ADJS)
• offset applied to Primary CPICH measurements before event evaluation
by the UE
404
Ec/Io
Reporting
P CPICH 1 Range
P CPICH 2
AdjsEcNoOffset
Enlarging Cell 3 by x dB
P CPICH 3
time
Reporting Reporting
Event 1B Event 1A
405
IFHO/ISHO Process Overview
HO
HOTriggering
TriggeringThresholds
Thresholdsset
setininRNC
RNC About 25 HO parameters
Event
EventTriggered
TriggeredCoverage/Capacity
Coverage/Capacity 5 Coverage/Capacity HO Reasons
based
based HO fulfilledininRNC
HO fulfilled RNC
(31 Intra-Freq neighbours)
RNC
RNCcommands
commandsselected
selectedUE(s)
UE(s)totostart
start 48 Inter-Freq neighbours
IF/IS measurements
IF/IS measurements 32 Inter-System neighbours can be measured
Measurements
Measurementsare
aredone
doneinin
Compressed
CompressedMode
Mode(CM)
(CM)
UE
UEreports
reportsGSM
GSMcells
cellswith
with Max 32 neighbours could be measured
strongest
strongest RSSI signals toRNC
RSSI signals to RNC
RSSI
RSSImeasurements
measurementsandandBSIC
BSIC Reporting cells are active set cells
verification for GSM cells
verification for GSM cells (max 3) + max 6 IFHO, max 6 ISHO neighb.
RNC
RNCmakes
makesHO
HOdecision
decisionand
and Different decision methods for IF HO
commands UE to target cell
commands UE to target cell Only one decision method for IS HO
406
IFHO/ISHO measurements- difference
IF-HO
• IF measurements Measurement Target Cell found
Trigger
WCDMA IF -measurements
IF-HO
Decision
• IS measurements
IS-HO
Measurement IS-HO
Trigger Target Cell found
Decision
WCDMA RSSI meas. BSIC verification
• BSIC verification required in case neighbour list includes multiple

GSM neighbours using the same RF carrier
407
IF/IS Handover Measurement Triggering Reasons
1. Low measured absolute 2 . Low measured absolute
CPICH Ec/No, event 1E/1F CPICH RSCP, events 1E/1F
FMCI: IFHOcauseCPICHEcNo FMCI: IFHOcauseCPICHrscp,
FMCG: GSMcauseCPICHEcNo FMCG: GSMcauseCPICHrscp
3. UE Tx power approaches
its maximum allowed power, 4. DL DPCH approaches its
Measurement maximum allowed power
event 6A/6D Trigger
FMCI: IFHOcauseTxPwrUL FMCI: IFHOcauseTxPwrDL
FMCG: GSMcauseTxPwrUL FMCG: GSMcauseTxPwrDL
5. Quality deterioration report 6 . Others

from UL outer loop PC - Load and Service based HO
FMCI: IFHOcauseUplinkQuality - IMSI based HO
FMCG: GSMcauseUplinkQuality - Emergency ISHO
Frequency Measuring Control for Inter-Frequency = FMCI

Frequency Measuring Control for Inter-System (GSM) = FMCG
408
1. Measurement trigger CPICH Ec/No
Reporting event: 1E: A P-CPICH exceeds an absolute threshold
1F: A P-CPICH falls below an absolute threshold (HO triggered if all)
e.g. P-CPICH Ec/No
Cell 1 Cell 2
Cell 3
absolute 1E: HHoEcNoCancel

threshold
1F: HHoEcNoThreshold
time
1E: HHoEcNoCancelTime 1F: HHoEcNoTimeHysteresis
409
1. Measurement trigger CPICH Ec/No
• RNC starts IF/IS measurement when event 1F occurs for all cells in the active
set: A Primary CPICH becomes less than an absolute threshold
• RNC stops IF/IS measurement when event 1E occurs for at least one cell of
the active set : A Primary CPICH becomes better than an absolute threshold
• Note:IF/IS measurements can be stopped if event 1Fs are cancelled by
events 1E only when IFHO/ISHO was not successful and only inside the
time between CM measurements, specified by the
time InterFreqMinMeasInterval GsmMinMeasInterval/default 10s,
recommendation 2s .
• Filtering applied before event evaluation in the UE:
– FMCS: EcNoFilterCoefficient/0= 200ms filtering period
410
2. Measurement trigger CPICH RSCP
• UE continually monitors pilot channels of BTSs in AS

• If RSCP of a Node B falls below threshold, HHoRscpThreshold, UE sends
event 1F report
• RNC starts IF/IS measurements when event 1F occurs for all cells in AS
• RNC stops IF/IS measurements when event 1E occurs for at least one cell of
AS
• Note:IF/IS measurements can be stopped if event 1Fs are cancelled by
events 1E only when IFHO/ISHO was not successful and only inside the
time between CM measurements, specified by the
time InterFreqMinMeasInterval GsmMinMeasInterval/default 10s,
recommendation 2s .
• UE filtering applied before event evaluation using HHoRscpFilterCoefficient
(FMCS) /200ms, range 200…1600ms
411
Reporting Events 6A, 6B and 6D
6A: The UE Tx power exceeds an absolute threshold
6B: The UE Tx power falls below an absolute threshold
6D: The UE Tx power reaches its maximum value (Not used in Nokia RAN)
UE Tx Power
6D
min(UEtxPowerMaxDPCH, P_MAX)
UE Transmitted Power
Tx Threshold
6A
6B
time
InterfreqUETxPwrTimeHyst time to
GSMUETxPwrTimeHyst trigger
412
3. Measurement trigger UE Tx Pw (UL Coverage)
• Event 6A parameters:
– Time-to-trigger: 0s
– UE Transmitted Power threshold = min (UEtxPowerMaxDPCH, Pmax) –
UE_TX_POWER_threshold
– UEtxPowerMaxDPCH is the maximum allowed UL tx power on DPCH in the
active cell (range -50..24 dBm, step 1, default 21 dBm), Pmax is the maximum
RF output power
– UE_TX_POWER_Threshold depends on the type of service:
• FMCI: InterfreqUETxPwrThrAMR/-3 dB (range -10..0, step1)
• FMCI: InterfreqUETxPwrThrCS/-3 dB (same range and step)
• FMCI: InterfreqUETxPwrThrNrtPS/-3 dB
• FMCI: InterfreqUETxPwrThrRtPS/-3 dB
• FMCG: GSMUETxPwrThrAMR/CS/NrtPS/RtPS -1/-3/-1/-3/ dB
• Event 6B parameters: time-to-trigger
– FMCI: InterfreqUETxPwrTimeHyst/1280 ms
– FMCG: GSMUETxPwrTimeHyst/1280 ms
• Event 6D parameter: time-to-trigger: 0s 413
4. Measurement trigger DL DPCH (DL Coverage)
• RNC orders UE to make IF/IS measurements when DL TX power of a single radio link
reaches threshold defined as:
DL_code_PWR - PowerOffsetDLdpcchPilot > Ptx_RL_max + DL_DPCH_TXPWR_Threshold
– DL_CODE_PWR = measured DL code power example: -100dBm > 33dBm+XXX+(-1dB)

on the pilot bits of the DPCCH-field
– Ptx_RL_max = CPICH Pwr + MAX_DL_DPCH_ TXPWR
• CPICH Pwr = Tx power of the CPICH of an active cell. PtxPrimaryCPICH (serving
RNC) or AdjsCPICHTxPwr (drifting RNC)/33 dBm, range -10…50 dBm, step 0.1
dBm
• MAX_DL_DPCH_TXPWR = maximum transmission power level of the DPDCH
symbols a base station can use on the DPCH, expressed as a relative value to the
CPICH power
– DL_DPCH_TXPWR_Threshold = depends on the service type
• FMCI: InterfreqDLTxPwrThrAMR/CS/RtPS/NrtPS -1/-3/-3/-1 dB (range -10 dB..0,
step 0.5 dB)
• FMCG: GSMDLTxPwrThrAMR/CS/RtPS/NrtPS -1/-3/-3/-1 dB (range -10 dB..0,
step 0.5 dB)
– In Addition, the DL data rate allocated to the user must not exceed
• WCEL: HHoMaxAllowedBitrateDL
414
NOTE: Either WCDMA or GSM will be measured, not both systems
5. Measurement trigger UL Quality
HHoMaxAllowedBitrateUL EnableULQualDetRep
RNC
RNC 1st Quality deterioration report from PC to HC inside RNC
2nd Quality deterioration report (while the condition is

Target SIR
satisfied the message is periodically repeated)

Max SIR target
∆SIR
∆SIR Actual SIR target
Min SIR target
default 5s time
ULQualDetRepThreshold (recommendation 2 s)
415
ISHO measurements
• After HO triggering message is sent to RNC, a RRC message ”Measurement

Control” is sent to UE containing details of the measurement that the UE
must execute. Measurement reporting is periodical. Max 6 GSM cells could
be measured by UE and reported to RNC. No filtering in measurements by
UE
RRC: ”Measurement control ” message
(GSM RSSI measurements)
RNC
UE
• Upon reception of the measurements reported by the UE, RNC applies a
sliding averaging window to the RXLEV measurements. The averaged levels
are used as input to the IS-HO decision algorithm.
RNC The first measurement report

UE has info from the best GSM cell:
BCCH freq & RSSI, no filtering
RRC: ”Measurement report” used in UE
GsmMeasRepInterval
(default 0.5s) RRC: ”Measurement report” RXLEV measurements are reported through
”Measurement report” messages
416
ISHO: BSIC Verification
• After the selection of the target GSM cell, the RNC sends to UE the RRC
message "Measurement control” which includes details to measure ”BSIC” .
• UE stops RSSI measurements and updates the transmission gap pattern to
the pattern used for BSIC decoding. The measurement reports are sent
periodically to RNC.
• If the UE is unable to decode the BSIC during the given period, the BSIC
measurement operation is aborted. IS-HO
IS-HO trigger Target Cell found command
WCDMA RSSI meas. BSIC verification

TRSSI depends on TBSIC depends on :
RRC: ”Measurement control ” message
(BSIC decoding) •Transmissiongap pattern for
•Numbers of GSM cells in
the BSIC decoding
RNC neighbor cell list
•Reporting interval
•Sliding averaging window
•Transmission gap pattern for
RSSI measurements
417
Compressed Mode (CM)
• CM was introduced to WCDMA to allow inter-frequency (system) Handovers
• CM is used to create idle periods (gaps) in the transmission during which ncell
measurements on another frequency can be made
Measurement
Measurementgap
gap
Power / Compressed
Data Rate Mode
Normal frame Normal frame Normal frame
• Gaps can be created using single or double frame approach (Nokia supports both)
• Because same data amount is sent in a shorter time more power is needed during
CM (both in UE and BTS)
=> affects WCDMA coverage
• Fast Power control information might be lost during the gap => higher Eb/No
=> affects WCDMA capacity
• Compressed frames may be lost if power control is not set correctly
=> affects WCDMA quality
• CM methods are Spreading Factor Halving and Higher Layer Scheduling
418
Measurement Gap
Length of gap:
3, 4, 5, 7 timeslots
Single-frame method
TGL
radio frame radio frame

Length of gap:
3, 4, 5, 7, 10, or 14 timeslots
Double-frame method
TG
L
radio frame radio frame
Min transmitted TSs per Frame: ≥ 8 419

IFHO Decision
It is not sufficient to have triggers, which detect that a Handover has to be done. Criteria are required to
decide, whether to carry out an intra-frequency, inter-frequency, or inter-RAT Handover – if there is
the possibility for a Handover at all!
Conditions shown here have to be satisfied, before an inter-frequency Handover can be conducted. The best
neighboring cell must fulfil following criteria („best“ according to Ec/No):
• Quality Criterion in case trigger was DL DPCH Power or CPICH Ec/No:

AVE_RSCP_NCELL (n) > AdjiMinRSCP(n) + max(0, AdjiTxPwrDPCH(n) - P_MAX)
Signal level from new cell > required Signal level in new cell + adjustment due to allowed power
AVE_EcNo_NCELL (n) > AVE_CPICH_EcNo + AdjiEcNoMargin(n)

EcNo on new cell better than EcNo in old cell + margin
• Pathloss Criterion in case trigger was UE Tx Power, CPICH RSCP or UL DPCH Quality:
AVE_EcNo_NCELL (n) > AdjiMinEcNo(n)
EcNo on new cell better than required EcNo in new cell (default –14 dB)
CPICH_POWER - AVE_CPICH_RSCP >

CPICH_POWER_NCELL (n) - AVE_RSCP_NCELL (n) + AdjiPlossMargin(n)
PathLoss in old cell > PathLoss in new cell + Margin (default 2 dB)
(continued)
420
ISHO Decision
The following conditions have to be satisfied, before an inter-system Handover can be conducted.
The best neighboring cell must fulfil following criterion („best“ according to Ec/No):
AVE_RXLEV_NCell(n) > AdjgRxLevMinHO(n) + max(0, AdjgTxPwrMaxTCH(n) - P_max)
where
• AVE_RXLEV _Ncell(n) is the averaged GSM carrier RSSI value of the GSM neighbor
cell(n), averaging done directly from dBm values (no linear averaging);
GSMMeasAveWindow / 6, 1…32 meas report.
• AdjgRxLevMinHO(n) determines the minimum Required RSSI (dBm) level of the
neighbor cell(n).
• AdjgTxPwrMaxTCH(n) indicated the maximum Tx power level (dBm) an UE may use in
GSM neighbor cell(n).
• P_MAX is the maximum UE power capability.
• If several GSM cells fulfil the equation above, cells are ranked based on HOPG:
AdjgPriorityCoverage/0, range 0…7
• A cell is ranked higher than another cell if it has a higher priority level even though
its signal strength condition was worse
• Inter-system HO could be forbidden during the first measurements reports from the UE, to let
the UE report all the candidate inter-system cells in its neighbour hood.
• FMCG NcellSearchPeriod / 0
421
ISHO 3G 2G - AMR Signalling Flow
UE Node B RNC CN
RRC: Measurement Control
ISHO triggering (3
RRC: Measurement Report reasons are possible)
NBAP: Radio Link Reconfiguration Prepare
NBAP: Radio Link Reconfiguration Ready
Initial Compressed
NBAP: Radio Link Reconfiguration Commit Mode Configuration
RRC: Physical Channel Reconfiguration
RRC: Physical Channel Reconfiguration Complete
NBAP: Compressed Mode Command
RRC: Measurement GSM RSSI
RRC: Measurement Report Control Measurement
RRC: Measurement Control GSM BSIC
RRC: Measurement Report Identification
RANAP: Relocation Required
RRC: Handover from UTRAN Command RANAP: Relocation Command

RANAP: IU Release Command
RANAP: IU Release Complete
• BSIC verification always performed for AMR calls – no interrupt in voice call
422
ISHO 3G 2G - PS Signalling Flow
UE Node B RNC CN
RRC: Measurement Control
RRC: Measurement Report ISHO triggering (5 reasons are possible)
NBAP: Radio Link Reconfiguration Prepare
NBAP: Radio Link Reconfiguration Ready

NBAP: Radio Link Reconfiguration Commit Initial Compressed
Mode Configuration
RRC: Physical Channel Reconfiguration
RRC: Physical Channel Reconfiguration Complete
RRC: Measurement Control GSM RSSI Measurement

RRC: Measurement Report
RRC: Cell Change Order from UTRAN

RANAP: SRNS Context Request
RANAP: SRNS Context Response
RANAP: SRNS Data Forward Command
RANAP: IU Release Command
RANAP: IU Release Complete
• In most cases BSIC verification is not required (data interrupt as UE moves to 2G)
• PS makes use of RRC: CELL CHANGE ORDER FROM UTRAN message
423
Load based Handover GSM WCDMA
Load of GSM cell

100% Load reason More
Speech
Speechto toWCDMA
WCDMA Morecapacity
capacity
Handovers only with high GSM load for speech
for speech
only with high GSM load
for speech
80%
Packet
PacketData
DatatotoWCDMA
WCDMA Higher
Higherbit
bitrates
rates
regardless
regardless of GSMload
of GSM load for data users
for data users
Packet
PacketBCCH
BCCHparameters
parameterscan
canbe
beused
used
to push GPRS mobiles to WCDMA
to push GPRS mobiles to WCDMA
0%
424
Load and Coverage Reason Handover
GSM high Load

Loadreason
reason
loaded Handover
Handoverfor
forspeech WCDMA
and for HSCSD
speech WCDMAcan canbe
beused
used
and for HSCSD to
torelieve
relieveGSM
GSMoverload
overload
WCDMA low
loaded GSM
GSMcan
canbe
beused
usedto
toextend
extend
WCDMA
WCDMAcoverage
coveragearea
area
GSM
GSM GSM
GSM GSM
GSM GSM
GSM
WCDMA
WCDMA WCDMA
WCDMA
GSM
GSM WCDMA
WCDMA GSM
GSM
Load reason Coverage reason

Handover Handover
Mobile moving 425
Load based Handover Process: GSM WCDMA
Handover
Parameters are sent from the BSC to the
HandoverTriggering
TriggeringThresholds
Thresholdsset
setininBSC
BSC mobile in the Measurement Information
message
In the S10.5 MS is measuring the UTRAN cells
Inter-RAT
Inter-RATmeasurements
measurementsstarts
startsinincase
case continuously if they are defined as a neighbor ,
The RXLEV
The RXLEV of the serving cell is aboveoror
of the serving cell is above
below
belowthe
thegiven thresholdQsearch_C
giventhreshold Qsearch_C The number of GSM neighbor reported could be 3-6
depending on the set value of parameter
FDD_MULTIRAT_REPORT (0,1,2=def,3).
Handover
HandoverDecision
Decisionisisdone
doneinincase
caseofof The load is measured in the cell level by
Load
Loadofofthe
theserving cell>>Load_Threshold
servingcell Load_Threshold comparing the occupied TCHs to the available
and CPICH Ec/No> min Ec/NoThreshold
and CPICH Ec/No> min Ec/No Threshold TCHs (20 s interval used)
Load Threshold for Speech Calls or

MS
MSselects
selectsthe
thetarget
targetUTRAN
UTRANcell
cellbased
based Transparent Data inter-RAT Handover:
on measurement results
on measurement results utranHoThScTpdc (Min Traffic Load TDHO),
default = 80%
Handover
Handovercommand
commandisissend
sendtotoMSC
MSC minEcnoThreshold (Min CPICH EC/NO
Threshold), default –15 dB
426
Cell Reselection 2G -> 3G
Cell Reselection L
ist
BCCH: FDD_Qmin
, FDD_Qoffset
GSM MS starts WCDMA measurements if :

at least every 5 s
RLA_C< F(Qsearch_I) for 0<Qsearch_I<=7
or
RLA_C> F(Qsearch_I) for 7<Qsearch_I<=15
check Measurement Updates:

• RLC_C value of the serving cell
• + least 6 strongest non serving GSM cells
Fdd_Qoffset: select
If, for suitable UMTS cell allways (value is -infinity)
no priorities & for a period of 5 s:

between CPICH RSCP > RLA_C + FDD_Qoffset WCDMA cell
WCDMA and reselection
neighbors
CPICH Ec/No ≥ FDD_Qmin FDD_Qmin= -13 dB
427
Cell Reselection 2G -> 3G
• Re-selection measurements are controlled by the parameter threshold to

search WCDMA RAN cells (QSRI)
– The parameter defines a threshold and also indicates whether these
measurements are performed when RLA_C (a running average of
received signal level) of the serving GSM cell is below or above the
threshold
GSM MS starts WCDMA measurements in case

running average signal (RLA_C) level is below or above
certain threshold:
RLA_C # Qsearch _I (Non GPRS)
428
Measurements for 2G 3G inter-system cell reselection
•Triggering of the 3G cell measurements

Depending on operator´s 2G – 3G interworking strategy parameter Q_search_I should
planned accordingly.
In the best case, 3G
In the best case, 3G cell measurements
cell measurements are restricted to the
are possible when condition:
RLA_C level < –74 RLA_C level > –78
dBm dBm
GSM
GSM 3G 3G 3G
GSM
Configuration 1 Configuration 2 Configuration 3

RLA_C< F(Qsearch_I) RLA_C> F(Qsearch_I) RLA_C< ∞
( 0<Qsearch_I<=6 ) ( 7<Qsearch_I<=15 ) (always).
(Qsearch_I=7)
If camping in a 3G cell is priority then:
- Configuration 1 is not acceptable because camping on 3G in zones with good GSM coverage is not possible.
- Configuration 2: camping on 3G is not possible in high medium&poor GSM coverage.
- Configuration 3: camping on GSM/3G is possible and not constraint to other system’s coverage.
Therefore, configuration 3 is the best choice at expenses of decreasing the battery lifetime of the terminal
429
Cell Re-selection Parameters
• Qsearch_I and Qsearch_P define the threshold for non-GPRS/GPRS

(respectively) capable UEs to measure 3G neighbour cells when a running
average of the received downlink signal level (RLA_C) of the serving cell is
below (0-7) or above (8-15) the threshold
Value 0 1 … 6 7 8 9 10 … 14 15
dBm -98 -94 … -74 Always -78 -74 -70 … -54 Never
If RLA_C < -94 UE starts 3G If RLA_C > -70 UE starts 3G

UE always measures 3G cells measurements
measurements
• FDD_Qoffset and FDD_GPRS_Offset the non-GPRS/GPRS (respectively) capable

UEs add this offset to the RLA_C of the GSM cells. After that the UE compares the
measured RSCP values of 3G cells with signal levels of the GSM cells
Value 0 1 2 3 … 8 … 14 15
dBm -always -28 -24 -20 … 0 … 24 28
Always select
Reselect in case RSCP > GSM
irrespective of RSCP 430
value RXLev (RLA_C) +28dB
Cell Re-selection Parameters
• FDD_Qmin, defines minimum Ec/No threshold that a 3G cell must exceed, in
order the UE makes a cell reselection from 2G to 3G. This is the new mapping
table
Fdd_Qmin mapping
Aif parameter 0 1 2 3 4 5 6 7
Fdd_Qmin (old) [dB] -20 -19 -18 -17 -16 -15 -14 -13
Fdd_Qmin (new) [dB] -20 -6 -18 -8 -16 -10 -14 -12
• FDD_REP_QUANT defines the reporting quantity for UTRAN cell (1=Ec/No or

0=RSCP, default 1). This parameter is hidden and it can not be tuned
• 3G_Search_PRIO parameter allows to extend the BSIC reporting interval from
10 s to 13s by checking the BSICs from non-serving BCCH carriers and also
make WCDMA measurements as often as possible. Range is 0=no, 1=yes,
default value .This parameter is hidden and it can not be tuned
• Defined in the SI2quater and PSI3quater if PBCCH is allocated
FDD_(GPRS) FDD_REP_
FDD_REP_ 3G_search_
Parameter
Parameter Qsearch_I
Qsearch_I Qsearch_P
Qsearch_P Qsearch_P
Qsearch_P FDD_QMin
FDD_QMin FDD_(GPRS)
Qoffset QUANT
3G_search_
PRIO
Qoffset QUANT PRIO
Name 2(3)quarter
Nameforfor 2quater 2quater 3quater
2quater 2quater 3quater 2ter
2ter 2quater
2quater 2ter
2ter 2(3)quarter
Sys Info (SI)
Sys Info (SI)
(PBCCH)
(PBCCH)
431
Cell Reselection 3G 2G
• Whilst camping in a 3G cell the UE performs intra-frequency, inter-frequency, and inter-system
measurements based on the measured CPICH EcNo of the serving cell according to the
following rules:
– Serving cell parameters Sintrasearch, Sintersearch and SsearchRAT are compared
with Squal (CPICH Ec/No - Qqualmin) in S-criteria for cell re-selection
– UE will measure neighbor cells depending on how parameters are set & sent (if
parameters are not sent UE shall measure all cells)
• 1 – None (Squal > Sintrasearch )
• 2 - WCDMA intra-frequency (Sintersearch < Squal ≤ Sintrasearch)
• 3 - WCDMA intra- and inter- frequency, no inter-RAT cells (SsearchRAT < Squal ≤ Sintersearch)
• 4 - WCDMA intra- and inter-frequency and inter-RAT cells (Squal ≤ SsearchRAT )
Sintrasearch Sintersearch SsearchRAT
4 3 2 1
WCDMA
CELL 432
Cell Reselection 3G 2G
CPICH EcNo UE starts GSM measurements if

CPICH Ec/No < qQualMin + sSearchRAT
Serving WCDMA cell
SintraSearch First ranking of all the cells based on calculation, with
CPICH RSCP (WCDMA) and RSSI (GSM) hysteresis parameter
SinterSearch Rs = CPICH RSCP + Qhyst1

Rn= Rxlev(n) - Qoffset1
SsearchRAT Neighbour WCDMA or GSM
cell calculation with offset
qQualMin
parameter
No Yes
Rn (GSM) > Rs (WCDMA)
And
Rxlev (GSM) >QrxlevMin
Second ranking only for WCDMA

Cell re-selection
cells based on CPICH Ec/No
to GSM
Rs = CPICH Ec/No + Qhyst2
Cell re-selection to
Rn=CPICH_Ec/No(n)-Qoffset2
WCDMA cell of highest
R value 433
434
435
436
437
438
439
Resource Manager
• The main function of RM is to allocate logical radio resources of BS
according to the channel request by the RRC layer for each radio
connection
• The RM is located in the RNC and it works in close co-operation with the
AC and the PS
• The actual input for resource allocation comes from the AC /PS and RM
informs the PS about the resource situation
• The RM is able to switch codes and code types for different reasons such
as soft handover and defragmentation of code tree.
• Manages the BS logical resources
– BS reports the available logical HW resources
• Maintains the code tree,
– Allocates the DL channelization codes, UL scrambling code, UL
channelization code type
• Allocates UTRAN Registration Area(URA) specific Radio Network
Temporary Identifier(RNTI) allocated for each connection and reallocated
when updating URA 440
Resource Manager
Spreading
• Spreading = channelization and scrambling operations (producing the
signal at the chip rate, i.e. spreads the signal to the wideband)
• Downlink: Scrambling code separates the cells and channelization
code separates connection
• Uplink: Scrambling code separates the MS's, channelization code
separates the DPDCHs in case of multicode
• The length of the channelization code is the spreading factor
• All physical channels are spread with channelization codes, Cm(n)
and subsequently by the scrambling code, CFSCR
• The code order, m and the code number, n designates each and
every channellization code in the layered orthogonal code
sequences.
user data widespread data
chanellization scrambling
code code 441
Spreading Code Allocation
• Code Allocation Algorithm chooses the correct spreading code depending on the
TFC type
C 3 (0)=(…)
C 2 (0)=(1,1,1,1)
C 3 (1)=(…)
C1 (0)=(1,1)
C 3 (2)=(…)
C 2 (1)=(1,1,-1,-1)
C 3 (3)=(…)
C 0 (0)=(1)
C 3 (4)=(…)
C 2 (2)=(1,-1,1,-1)
C 3 (5)=(…)
C1 (1)=(1,-1)
C 3 (6)=(…)
C 2 (3)=(1,-1,-1,1)
C 3 (7)=(…)
Code Order Code Order 1 Code Order 2 Code Order 3
0 (SF 1) (SF 2) (SF 4) (SF 8)
The codes are layered from 0 to 11 according to the Spreading Factor (SF)
• Cm(n) : The code order, m, and the code number, n, designates each and every
code in the layered orthogonal code sequences
• In DL code order 2 to 8 (SF 4 to 512) are available (Nokia RAN does not support
SF = 512)
• In UL code order 2 to 7 (SF 4 to 256) are available 442
Spreading Code Allocation – Example
Ordinary downlink speech channel, requires 30 ksps physical capacity (AMR
12.2 – 4.75 kbit/s). The code order is 8, which means there are 128 chips to
illustrate one symbol (2n, n=7). If the requested channel is 120kbit/s (including
DPCCH and channel coding) data channel, then the type of code is 6.
Spreading code cycle depends on the symbol rate.
Code Channel Channel

DPDCH
Order symbolrate bitrate SF
bits/frame
(ksps) (kbps)
9 15 7.5 512 40
8 30 15 256 160
7 60 30 128 340 AMR Speech

6 120 60 64 600
5 240 120 32 1400 64 kbps data

4 480 240 16 2880 128 kbps data
3 960 480 8 6080 384 kbps data
2 1920 960 4 12480
443
Downlink Code Allocation
•HSDPA with 5 codes allocated at cell MAC-hs start-up when HSDPA is enabled
•Code allocation is dynamic in future releases when more than 5 codes are allocated
SF = 1
SF = 2
SF = 4
SF = 8
SF = 16
SF = 32 Codes for 5
SF = 64 HS-PDSCH's
SF = 128
SF = 256
Code for one
HS-SCCH
Codes for the cell common channels
•166 codes @ SF=256 available for the associated DCHs and non-HSDPA uses
444
Downlink Code Allocation – Common Channels
• Pilot (P-CPICH) and BCCH (P-CCPCH) need a fixed code allocation
– P-CPICH: CH256,0
– P-CCPCH: CH256,1
• AICH and PICH codes in Nokia RAN
– AICH: CH256,2
– PICH: CH256,3
• S-CCPCH code allocation depends on number of active S-CCPCH in Nokia
RAN
– With 1 S-CCPCH: CH64,1
SF = 32
SF = 64 S-CCPCH
SF = 128 X
SF = 256 X X
P-CCPCH
P-CPICH
AICH
PICH
445
Spreading Code Allocation
• A code is always allocated from the optimum location in the code tree. It
makes the allocated code and the codes in the branches below and above
the allocated code unavailable
• Code tree will fragment quickly if releases is not re-arranged
• Re-arrangements in the code tree is done by reallocating the codes in better
locations
• The above code tree has 4 codes of equal order. The best locations are in
the same branch and very close to one another. The badly located codes are
released and optimally reallocated allowing the use of upper layer codes
• Codes are only reallocated when there is a benefit at two code tree layers
above the code being reallocated
446
Part XIII
Drive Test Analysis
447
Drive Test Process
Cluster Data Preparation Data Analysis Trouble

Preparation and Collection • Levels of Shooting
• Define Cluster • Services to test Reporting • Root Cause
• Define Drive Route • KPI definitions • Failure Type Analysis
• Equipment Breakdown • Corrective Action
• Exclusions • Post Processing • Further data
• Call Patterns functionality logging (maybe)
• OSS Alarms • KPI Gap Analysis
• Network Stats
Why Drive the network?
• New Site Integration

• RF Tuning
• Network Optimisation
• Network Benchmarking (Golden Routes)
448
Drive Test Analysis – Test Equipment
• Nemo Outdoor Multi with UE’s, Scanner and GPS
Land Unit
NQMP
FTP
Server
• Important to drive with Call logging equipment and

3G and 2G scanner in the same vehicle
449
Drive Test Analysis – Test Equipment, Scanner
• The purpose of using the RF scanner is to be able to scan and

measure all used carriers/cells and their corresponding DL
scrambling codes.
– For low coverage areas

– For antenna installation problems
– For missing neighbours
– For coverage optimisation
– With the scanner you can get the following info from the surrounding
cells:
• Different Scrambling codes
• CPCIH RSCP value (dBm)
• CPICH EcNo value (dB)
450
Drive Test Analysis – Test Equipment, UE
• The Scanner measures all SCs, whereas the UE only measures SC
signals from the cells that the system has informed/ordered the UE through
the BCH (neighbour list) or via the “measurement control” message.
– RSCP Active/Monitored Set
– Carrier RSSI
– Ec/No Active/Monitored Set
– UL/DL Data Throughput
– The BLER downlink
– Pilot BER
– Random Access Initial Tx Power
– Random Access Preamble Count
– Random Access Preamble Step
– Random Access Tx Power
– SIR target (UE dependent)
– UE Tx Power
– Call Statistic: AMR, CS and PS data calls
451
Drive Test Analysis – Call Patterns
• Enough call samples have to be made to make the measurement statistically valid.
• In a 50 call sample one dropped call will cause a change in performance of -2%
• In a 500 call sample one dropped call will cause a change in performance of -0.2%
• Call length should be defined at the beginning
• We can use different call testing patterns for different optimisation techniques
• Short Calls (for Calls setup performance and delay)
• Long calls (for Drop call performance and SHO performance)
AMR Call PS Call

Mobile Originated Calls (MOC) • GPRS Attach,
• 2 min calls • PDP Context Activation
• 30 sec idle • FTP Download (1MB file)/FTP Upload (500 KB
file)
• UE in Dual mode (2G/3G) • PDP Context Deactivation
Mobile Terminated Calls (MTC) • GPRS Detach
• 2 min calls • Alternate download and upload with 30 sec idle
• 30 sec idle time
• UE in dual mode (2G/3G) • Session is upload or download
• UE in Dual mode (2G/3G) 452
PS Data KPI Report
Drive Test Analysis – Defining KPIs

KPI Thresholds Value Unit
Attach Time 4 sec
PDP Activation Time 2 sec
AMR Statistics FTP Download Throughput 110 kbit/s
FTP Upload Throughput 55 kbits
Threshold KPIs
MOC Setup time 5 sec Signalling Statistics Count Success rate
MOC CCR 99 % Attach Attempt 155
Attach Success 140 90.32%
Attach Time more than threshold 5 3.57%
RAW END USER Attach Time less than threshold 135 96.43%
Event Count Ratio Count Ratio Attach Failed 15 9.68%
Call Attempts 132 122 Average Attach Setup Time 1.36
Activate PDP Context Attempt 124
Call Setup Success Rate 108 81.8% 108 88.5%
Activate PDP Context Success 124 100.00%
Call Setup Failure Rate 24 18.2% 14 11.5% Activation Time more than threshold 2 1.60%
Failures due to Tool (TSF) 10 41.7% Activation Time less than threshold 123 98.40%
Failures due to Core Problem 10 41.7% 10 71.4% Activate PDP Context Failed 0 0.00%
Failure ASU (Sync) Problem 2 8.3% 2 14.3% Average PDP Context Activation Time 0.96
FTP Download Attempts 51
Failure due to Low Coverage Levels 2 8.3% 2 14.3%
FTP Download Success 48 94.12%
Call Setup Success and Setup Time > 5s 7 6.5% 7 6.5% FTP Download throughput more than threshold 25 52.08%
Long Setup due to slow cell reselection 0.0% 0 0.0% FTP Download throughput less than threshold 23 47.92%
Long Setup due to clash with InterRAT reselection 0.0% 0 0.0% Average FTP Download Throughput 107.02
Long Setup due to Unknown (suspect UE) 0.0% 0 0.0% FTP Upload Attempts 32
FTP Upload Success 30 93.75%
Long Setup due to Unknown 0.0% 0 0.0% FTP Upload throughput more than threshold 20 66.67%
Average Call Setup Time 3.66 3.66 FTP Upload throughput less than threshold 10 33.33%
Call Completetion Rate 105 97.2% 105 97.2% Average FTP Upload Throughput 55.53
Call Drop Rate 3 2.8% 3 2.8% Data Transfer Cut-off Ratio 6.02%
PDP Context Dropped 4 3.23%
Call Drop Poor 3G Coverage 1 33.3% 1 33.3% Deactivate PDP Context Request 121
Call Drop on GSM due to Interference 2 66.7% 2 66.7% Deactivate PDP Context Accept 121 100.00%
Overall Call Completion Rate 105 79.5% 105 86.1% Deactivate PDP Context Failure 0 0.00%
ISHO Attempt 14 14 Detach Request 281
Detach Accept 129 45.91%
ISHO Success 14 100.0% 14 100.0%
Overall Data Session Completion Rate 78 62.90%
ISHO Failed 0 0.0% 0 0.0%
ISHO Failed cause physical channel failure 0 N/A 0 N/A RAU Statistics Count Success rate
RAU Attempt 22
• Need to agree RAU Success

RAU Failed
13
9
59.09%
40.91%
•What ‘raw’ figures will contain Cell Reselection Statistics

Cell Reselection from UMTS to GSM Attempts
Count
2
Success rate
Cell Reselection from UMTS to GSM Success 1 50.00%
•What ‘End User’ will contain Cell Reselection from UMTS to GSM Delay E2E
Cell Reselection from UMTS to GSM Delay Signalling
15.27
N/A
Cell Reselection from GSM to UMTS Attempts 1
Cell Reselection from GSM to UMTS Success 0 0.00%
•Other cuts of Data Cell Reselection from GSM to UMTS Delay E2E N/A
Time in System Seconds

453
Ratio
GSM 171 1.02%
Company Confidential UMTS 16559 98.98%
Drive Test Analysis – Test Case definition Examples
Short calls to test
Call Setup
Success
KPI ID Criteria Test Description Trigger Statistics Success
Rate
KPI-01 Voice 12.2 •1 MOC+ 1 MTC in the same van, •Call set-up trigger point: •Cluster_AMR_Overall_CCSR_% = Call 98%
kbps AMR 2 Ues •Party A’s UE sends 1st ‘RRC Connection Request’ Completion / Call Attempts *100
CCSR •1 call (party A to party B, party A •Call completion trigger point:
terminates the call): 15 seconds •Party A’s UE receives ‘RRC Connection Release (cause normal)’
call + 15 seconds idle. Party A to •Successful completion if:
be logged •15 sec call duration (from RRC Conn. Request to RRC Conn.
•350 calls Release)
•Party A’s UE receives ‘RRC Connection Release (cause normal)’
KPI-02 Voice DCR •1 MOC to PSTN, 1 Ue •Call Set-up trigger point: •Cluster_AMR_Overall_DCR_% = 1 - Call < 2%
•1 call: 90 seconds call + 15 •3G UE receives ‘RRC: Downlink Direct Transfer (Alerting for NZ, Completion / Set up Calls *100
seconds idle Connect Acknowledge for AU)’
•350 calls •Call completion trigger point:
•3G UE receives ‘RRC Connection Release (cause normal)’
•Successful completion if:
Long calls to test •90 sec call duration (from RRC Conn. Request to RRC Conn.
Drop call ratio Release)
•3G UE receives ‘RRC Connection Release (cause normal)’
KPI-03 Voice 3G- •1 MOC to PSTN, 1 Ue in dual •3G to 2G HO start point: •Cluster_Voice_3Gto2G_HO_Overall_SR_% = HO 98%
2G mode •UE receives ‘RRC: Handover From UTRAN Command’ Completion / HO Attempts *100
Handover •1 call: continuous call until 3G2G •3G to 2G HO completion trigger point:
(inter- & coverage border is passed •UE sends ‘Handover Complete’ to 2G BSS
intra-MSC) •50 occurrences as a minimum •Successful completion if:
•UE sends ‘Handover Complete’ to 2G BSS
…
KPI-08 PS PDP •Same as 10 •PS PDP activation trigger point: •Cluster_PDP_Context_Activation_SR_% = PDP 98%
context •350 occurrences •UE not PS attached Context Activation Completion / PDP Context
activation •UE sends 1st ‘RRC Connection Request’ Activation Attempts *100
•PS PDP activation completion trigger point:
•UE receives ‘RRC: Downlink Direct Transfer (SM: Activate PDP
Context Accept)’
•Successful completion if:
•UE receives ‘RRC: Downlink Direct Transfer (SM: Activate PDP
Context Accept)’
KPI-09 PS •Repeated FTP calls, 1 Mbyte file •Data Session Set-up trigger point: •Cluster_PS_DL_Throughput = Average 200 Kbps
throughput download, 1 Ue •UE receives 1st DL packet Throughput Over Data Sessions
downlink •384 kbps •Data Session Completion trigger point:
•Average throughput •UE receives last packet
•350 calls •Session output: 454
•Average downlink throughput
Need to Define the KPI measurement (from Drive test)
• Call Setup Success - CSSR (voice, circuit switched data)
– Successful call setup means that “DL/UL Direct Transfer (CC: Alerting)”
message is received by UE.
• Call Setup Time (voice, circuit switched)
– Call setup delay is measured from L3 messages, starting from “RRC
Connection Setup” message to “DL Direct Transfer (CC: Alerting)” message.
• Call Drop (voice, circuit switched)
– A dropped call occurs. The call will be dropped in case RRC connection release
(not normal release) message has been send from RNC to UE.
• Session Setup Success (packet switched)

– This is related to PDP context activation. Successfully activated PDP context
means that activate PDP context accept message has been sent from RNC to
UE (RRC: downlink direct transfer (SM:activate PDP context Accept)).
• Session Setup Time (packet switched)
– The “session setup time” is the delay between the time the UE sends the data
session activation request until GPRS attach and PDP context activation has
been successfully completed.
• Session Drop (packet switched)
– Session drop rate can be defined as the number of successful PDP
deactivations against number of successful PDP activations.
455
Definition of Call Set-Up Success Rate (CSSR)
UE RNC CN
BTS
RRC: Connection Request
RRC Connection Setup phase
Resource Reservation in RNC, BTS, Transmission
RRC: RRC Connection Request Setup
RRC
RRC Connection
Connection Access
Access phase
phase
RNC
RNC waits
waits reply
reply from
from UE
UE
RRC: RRC Connection Completed
RRC: Initial Direct Transfer cm service request
RANAP: Initial UE Message
DIRECT TRANSFER (Call Proceeding) Call Set-
up
RANAP: RAB Assignment Request
Success
RAB Connection Setup phase Rate
Resource Reservation in RNC, BTS, Transmission
Call Set-
RRC: Radio Bearer Set-up up Time
RAB
RAB Connection
Connection Access
Access phase
phase
RNC
RNC waits
waits reply
reply from
from UE
UE
RRC: Radio Bearer Setup Complete

RANAP: RAB Assignment Response
DIRECT TRANSFER (Alerting)
DIRECT TRANSFER (Connect)
DIRECT TRANSFER (Connect Acknowledge)
Successful call setup means
that “DL/UL Direct Transfer
(CC: Alerting)” message
456is
received by UE
Definition of the Call Completion Rate (CCR)
UE Node B RNC MGW
DIRECT TRANSFER (Alerting)

DIRECT TRANSFER (Connect)
DIRECT TRANSFER (Connect Acknowledge)
Call Established
Call Drop
Direct Transfer (Disconnect)
Rate
Direct Transfer (Release)
Direct Transfer (Release Complete)
Iu Release Command Call
Iu Release Complete
Duration
RRC Connection Release
RRC Connection Release Complete
Radio Link Deletion Request
Radio Link Deletion Response
ALCAP: Release Request
ALCAP: Release Response
ALCAP: Release Request
ALCAP: Release Response
Call Released
457
Definition of Session Setup Time - PS
UE WBTS RNC
CN
UE already has an RRC connection
INITIAL DIRECT TRANSFER (Attach Request)
INITIAL UE MESSAGE (GPRS Attach)
SCCP: Connection Request
SCCP: Connection Confirm

Downlink Direct Transfer (Authentication & Ciphering Request) RANAP: Authentication & Ciphering Request
Uplink Direct Transfer (Authentication & Ciphering Response) RANAP: Authentication & Ciphering Response
Security Mode Command RANAP: Security Mode Command
Security Mode Command RANAP: Security Mode Command

RANAP: Common ID
Downlink Direct Transfer: Identity Request RANAP: Identity Request
Uplink Direct Transfer: Identity Response RANAP: Identity Response
Downlink Direct Transfer: Attach Accept Direct Transfer: Attach Accept Session
Set-up
Uplink Direct Transfer: Attach Complete Direct Transfer: Attach Complete Time
INITIAL DIRECT TRANSFER (Active PDP Context Request) DIRECT TRANSFER (Active PDP Context Request)
RANAP: RAB ASSIGNMENT REQUEST
RRC: Radio Bearer Set-up
RRC: Radio Bearer Set-up Complete
RANAP: RAB ASSIGNMENT
RRC: Measurement Control RESPONSE
DIRECT TRANSFER (Active PDP context Accept)
458
Failure Breakdown
• Non-genuine failures
– Measurement system fault
It is beneficial to (Collection Tool or Analysis)
categorise call failures • Genuine failures
during the analysis and – RF issue (Coverage /
reporting Interference / Poor
dominance)
– Missing neighbour
– System issue WBTS
– System issue RNC
– Core network issue
UE
– System
– (Unknown )
Iub Iu
WBTS RNC Core NW
459
Failure Breakdown
• The KPI measurement conditions should be used to define exclusion

areas during drive test for acceptance of a cluster.
• All failures happening in those area, that do not respect the coverage
requirements should be discarded.
RSCP
Example of Call Success Criteria Ec/No
Signal Call
above Signal above
OK
Call Call -(B) Ec/No
RSCP Call
threshold
-80dBm OK
NOK -5dB threshold
µ
- -95dBm --12dB
Signal Signal
below below
RSCP Call Ec/No
OK
threshold -110dBm µ -20dB threshold
Threshold Call
Call-(A) NOK Discard sample
Time
460
Failure Breakdown
AMR Call Setup Failures
Call Set-up Failures Breakdown - Benchmark Drive
Call Setup Failure Breakdown - Final Drive
5% 5% Equipment 0%
0%
0%
Registration Clash
0% Equipment
15%
Core Network 19% 19% Registration Clash
20%
Fail in 2G Core Network
Fail in 2G
RAN (fixed in next release) 0%
RAN (fixed in next release)
Site maintenance (TMUK) 0% 0%
Site maintenance (TMUK)
0% RAN (Other)
RAN (Other)
10%
5% Missing neighbours
Missing neighbours
0% 3G RF 3G RF
0%
0%
slow cell Reselection 0% Slow cell reselection
10% site configuration & maintenance

site configuration & maintenance
(Nokia)
35% (Nokia) Other
Other
5%
52%
Call Set-up Failures Call Set-up Failures

Failure Type total % Failure Type total %
Equipment 1 5.0 Equipment 0 0.0
Registration Clash 4 20.0 Registration Clash 0 0.0
Core Network 0 0.0 Core Network 0 0.0
Fail in 2G 1 5.0 Fail in 2G 6 19.4
RAN (fixed in next release) 0 0.0 RAN (fixed in next release) 0 0.0
Site maintenance (TMUK) 0 0.0 Site maintenance (TMUK) 16 51.6
RAN (Other) 2 10.0 RAN (Other) 0 0.0
Missing neighbours 1 5.0 Missing neighbours 0 0.0
3G RF 7 35.0 3G RF 3 9.7
slow cell Reselection 3 15.0 Slow cell reselection 0 0.0
site configuration & maintenance (Nokia) 0 0.0
site configuration & maintenance (Nokia) 0 0.0
Other 1 5.0
Other 6 19.4
461
Drive Survey Analysis Process Summary Diagram
462
Non Genuine Call Setup Failure Scenarios
• Measurement systems are often not perfect and may introduce errors
in data collection or analysis
• Examples of non-genuine failures seen:
– Uplink: CM Service Abort within milliseconds from CM Service Request

– Call attempt during Location Area update (‘LA clash’)
– User initiated “UL CC Disconnect”
– Location Area update interpreted as call setup failure
– Cell reselection back to 3G from 2G interpreted as call setup failure
463
Non Genuine Call Setup Failures
• Measurement system failures by drive test tool

– “CM Service Abort” within milliseconds from “CM Service
Request”
– No time for response from NW
464
• Measurement system failures by
drive test tool
• Call attempt during Location
Area update (LA update clash)
LA Update
Request
Call attempt
RRC
Release
465
Measurement system failures by Actix workbook
– Successful or failed Location Area update interpreted as call setup failure
LA Update
Request
RRC
Connection for
Inter-RAT cell
reselection
466
Non Genuine Drop Call Scenarios
Measurement system failure examples:

– Drive test tool
• User initiated “UL CC Disconnect”
– Analysis Workbook
• Inter-Rat cell reselection from 2G to 3G interpreted as drop call
• Complete (e.g 90 seconds) call on either 3G or 2G
• No drop in the log file / same drop listed twice
467
Genuine Call Setup Failure Scenarios
– RF issue
• Interference / Dominance / Coverage
• Missing neighbour
– System Issue - BTS
• No response to “RRC Connection Request”
• “RRC Connection Reject” to “RRC Connection Request”
– System issue - RNC
• “CC Disconnect” after “Call Proceeding” due to “DL RRC Connection
Release”
– Core NW
• “CM Service Abort” after “CM Service Request”
– System issue (test number)
• “CC Disconnect” after “CC Progress”
468
Genuine Drop Call scenarios
• RF issue
– Interference / Dominance / Coverage
– Missing Neighbours
• System issue BTS
– Sudden “CC Disconnect” due to “DL RRC Connection Release”
– Sudden drop to idle, no disconnect messaging
• System issue RNC
– Sudden “CC Disconnect” due to “DL RRC Connection Release”
469
Failure Location
• Analyse the
signalling
flow to find
the location
of failure and
potential
cause
• UE log may
only capture
some of the
messages
470
AMR CS Call Phases
UE WBTS DRNC SRNC CN
1. Cell search, BCCH
decoding & RACH access
2. RRC connection set-up
3. UE <--> CS-CN signalling
4. RAB set-up
5. UE <--> CS-CN Signaling
6. Service Established
7. Branch addition/deletion & Active set update
8. Service Released
471
Call Setup Failure Start
Analysis Process A Coverage Optimization

Best server’s No
RSCP > -102dBm
Missing No
Neighbour ?
Dominance Optimization
Yes
Best server’s No
Ec/No > -12dB Yes Neighbour list Optimization
Yes
UL coverage & RACH
B parameter. Optimization
(changing serving cell)
No AICH(ACK) received? Report & Finish
(Check failure cause)
Yes No
“RRC Connection
Setup” received?
No “RRC Setup
Reject” received?
C
Yes Yes
AC optimization (check PrxNoise
D Report & Finish

(Reason of problem: L1 sync fail)
No (DCH) “RRC Connection setup
Completed” sent from UE?
& interferer around BTS)
Report & Finish

Yes
Report & Finish No “Radio Bearer setup

(Check failure cause) failure”Received?
Ye
s
E
Report & Finish
Check failure cause
(Not radio problem/cell update) 472
Call setup failures – RF issue A
• RF issue? Coverage / Interference / Dominance
See the example in Module 3 – RF Optimisation

473
Call setup failures – Missing Neighbour A
• Missing neighbour analysis over the whole route (3G-3G, 3G-2G)
• Search for failures due to missing 3G-3G neighbours
• Search for failures due to missing 3G –2G neighbours
– It is suggested to place 2G scanner to the test vehicle
474
Call Setup Failure Analysis- Block B - B
• The purpose of this activity is to check the Random Access Process
is working adequately by investigating whether AI (Acquisition
Indicator) has been received through DL AICH
• If AICH was not received by UE, the cause of the problem can be
classified into:
– Inadequate RAN parameter related to Random Access: RAN parameter
settings for pre-amble transmission or open loop power control
information is not correct.
– UL Coverage limit: UL coverage of UE is smaller compared to serving
cells DL coverage so that UE’s Tx power cannot reach serving cell.
• The Basic theory for RACH setup procedure and planning parameters
can be found in Module 6 – Parameter Optimisation
475
Call Setup Failure Analysis- Block B - B
UE WBTS RNC
Preamble/RACH
Acquisition Indicator/AICH
RRC: RRC Connection Request/PRACH

NBAP: RADIO LINK SETUP REQUEST
NBAP: RADIO LINK SETUP RESPONSE
RRC: RRC CONNECTION SETUP/FACH
L1 Synchronisation
NBAP: SYNCHRONISATION INDICATOR
RRC: RRC CONNECTION SETUP COMPLETE/DCH
UE in CELL_DCH state
476
RACH Process
B
L1 ACK / AICH
Downlink Not detected
BS
UEtxPowerMaxPRACH
PowerRampStepPRACHpreamble
…… ……
Uplink
MS Preamble Preamble
1 2
Message part
PowerOffsetLastPreamblePRACHmessage
PRACH_preamble_retrans
# PRACH preambles transmitted during one PRACH cycle without receiving AICH response
RACH_tx_Max
# preamble power ramping cycles that can be done before RACH transmission failure is reported
Initial preample power:

•Ptx = CPICHtransmissionPower-RSCP(CPICH) +RSSI(BS) +
PRACHRequiredReceivedCI
477
Call Setup Failure Analysis- Block B-
• Solutions for RACH optimisation
B
No
Max UE Tx power hit To
Toincrease
increasePRACH_Preamble_retrans
PRACH_Preamble_retrans
the UE_P_MAX(24dBm)? Or PowerRampStepPRACHPreamble
Or PowerRampStepPRACHPreamble
Yes
Yes
Is UL Interference Report
Reportthere
theremight
mightbebean
aninterfering
interfering
abnormally HIGH? source
source Nearby the servingcell
Nearby the serving cell
No
Change
Changethe
theServing
Servingcell
cellto
tocover
coverthe
theproblem
problemArea
Area
=> UE is too far to reach the serving cell
=> UE is too far to reach the serving cell
478
Call Setup Failure Analysis- Block B B
Open loop Power

Control parameters from
RACH Info message
479
Call setup failures – System issue BTS C
• No response to “RRC Connection Request”
– Good RF conditions
– Wrong MHA settings or cable loss settings can cause the site not to “hear” the
UE
– PrxNoise statistics, receive link parameters and HW units to be checked (faulty
MHA, wrong MHA parameters, wrong cable / feeder loss parameters, faulty
units)
480
Call setup failures – System issue BTS C
• “RRC Connection Reject” after “RRC Connection Request”
– Good RF conditions
– Admission Control can reject too many (or admit too many) connection
requests due to wrong PrxNoise measurements.
– PrxNoise statistics, receive link parameters and HW units to be
checked
481
Call Setup Failure Analysis
• UE has the appropriate DL/UL coverage but if RNC does not allow to
C
set up the RRC connection of the requested RAB (Radio Access
Bearer), Call setup will fail.
• Admission Control (AC) is involved in RRC connection setup. AC can
reject RRC reject RRC connection Setup due the DL Load, UL load or
DL Spreading codes
– Marginal Load Area:

• If measured UL (PrxTotal) or DL (PtxTotal) load exceeds target
thresholds (PrxTarget and PtxTarget) AC can still admit new RAB to
the cell if a new non-controllable load keeps below target thresholds
(in practice this means that AC can admit only new controllable load
RABs i.e. NRT RABs)
– Overload Area:
• If measured UL (PrxTotal) or DL (PtxTotal) load exceeds overload
thresholds (PrxTarget + PrxOffset and PtxTarget + PtxOffset) then
AC can't admit more RABs to the cell 482
Call Setup Failure Analysis C
• During the pre-optimization phase it is unlikely that AC will stop an RRC
connection setup during the drive testing because there are normally very
few UEs in the network. (Traffic loading is trivial)
• However, it should be checked that measured PtxTotal and PrxTotal are
less than PtxTarget (e.g. 40dBm) and PrxTarget (e.g. 4dB, 60% loading)
respectively.
• If DL AC does not allow RRC setup check the Tx power of WBTS, # of
channels transmitted, Signaling messages.
• If UL AC does not allow RRC setup: Check out if there is an interfering
source nearby the serving cell.
483
Call Setup Failure Analysis D
To check if Layer 1 Synchronization (slot/frame sync) has failed
• If “RRC Connection Setup” was received by UE but UE does not send “RRC
Connection Setup Completed”, we will report “L1 synchronization failure” and
have to check L1 system messages.
484
Call setup failures – System issue RNC E
• “CC Disconnect” after
“Call Proceeding”
• Good RF conditions
• Failures in RAB setup
occur between the “RAB
Assignment Request”
being received from Core
Network and the RAN
sending out Radio
Bearer Setup. Therefore
the failure is between
BTS and Core Network.
485
Call setup failures – System issue RNC E
• “CC Disconnect” after “Call
Proceeding” (cont.)
• An example (site shows high
values on counter
“RAB_STP_FAIL_CS_VOICE
_BTS” during the drive test
• In the recent check the
counter showed no failures.
486
Call setup failures – Core NW E
• “CM Service Abort” after
“CM Service Request”
• Good RF conditions
• “Security Mode
Command”-message not
received by UE, thus the
failure is believed to be at
Core Network. UE Node B RNC MGW
RRC Connection Establishment
Initial Direct Transfer (CM Service Request) SCCP: Connection Request
SCCP: Connection Confirm

Location Reporting Control
Common ID
Security Mode Command
• RRC: Initial Direct Transfer message is sent using acknowledged mode RLC to the CS core
domain. Routing is to be based upon the local P-TMSI
• The NAS message is not read by the RNC but is forwarded to the multimedia gateway. The NAS
message includes the IMSI as a UE identity
• The SCCP: Connection Request message establishes the connection orientated signalling link in
the same way as it was for the RRC connection phase.This does not reserve any resources for the
AMR call itself.
• The Connection Confirm message identifies the RNC with a destination local reference which is the
same as the source reference within the Connection Request message
• The Connection Confirm message identifies the CS core with a source local reference
• The CS core sends a RANAP: Location Reporting Control message to the RNC requesting
information regarding the location of a particular UE
• The RANAP: Common ID message specifies the IMSI belonging to the UE
487
• The Security Mode Command message triggers the start or stop of ciphering and integrity
protection.
Call setup failures – System Issue (test number)
E
• “CC Disconnect” after “CC Progress”
• Cause: recovery on timer expiry
• The call goes via IN SCP to a
recording.
• A static test was done by Nokia
Customer Care and in few instances
the call dropped after 30 seconds of
recording passed. Hence the problem
is associated with the test number not
the RAN
30
sec
Cause: recovery
on timer expiry
488
Call Drop Failure Analysis Process Start
SHO Failure Yes

SHO Failed
Analysis
No
ISHO Failure Yes

ISHO Failed
Analysis No
No
Best server’s No
Coverage Optimization
RSCP > -102dBm
Best server’s Missing No

Dominance Optimization
Ec/No > -12dB Neighbour
A Yes Neighbour list

Yes Optimization
Investigate
B possible BTS
or RNC
problem 489
Call Drop Failure Analysis Process (SHO Analysis)
Start
Yes
Check Iur Inter RNC HO
Yes
Check neighbour definition parameters
No
Yes No DL ASU
Fix SC Clash SC Clash
received
C
Yes
No
Yes Congestion on
Load Optimisation
target cell
UE Tx Power Max Check RF Levels
No
DL Tx Power Max No
Yes
D
Yes Load Optimisation/
Uplink Interference
External Interferer
Yes
No
Yes
Link Unbalanced CPICH Optimisation
490
Drop call failures – RF issue A
• RF drops mostly due to poor
dominance or interference
• Poor coverage could lead to ISHO,
although poor dominance or
interference can cause ISHO to
fail.
• Rapid field drop can cause drop
due to coverage
• Poor dominance or interference
can cause Compressed Mode
(CM) to start even if RSCP is still
good.
• In CM UE transmits with higher
power (more interference) and
spends less time on 3G (less
Poor dominance
accurate measurement reporting) causes Active Set
• Poor dominance or interference update failures
can lead to Active Set update
failures and eventually to drop call. 491
DL synchronisation is lost ->

UE has stopped transmitting
TrChAgg and DL
DPCCH BER
high
492
Fairly good
CPICH Pilot
EcNo
Transport Channel
BER. Btw UE<-
>RNC (MAC layer)
Sometimes DPCCH BER (btw UE<->WBTS)

can be a better indicator of what's happening
to the dedicated channel than the CPICH
EcNo, in particular in the case that power
control may not be tracking well
493
Drop call failures – System issue BTS B
• Sudden drop to idle, no disconnect messaging
– Site malfunctions to be checked
– In the example below site had faulty unit (WTR)
Drop to IDLE
494
Drop call failures – System issue RNC
B
• “CC Disconnect” due to DPCCH BER
“DL RRC Connection
Release”
• No response to UL
Measurement Reports
• In the example site had
no alarms, good RF &
BER
• Not able to add SC265
to Active Set, next call
on the same cell => no
failure.
• Difficult to troubleshoot if
the failure does not
happen systematically Sudden “RRC
=> follow up in the next Connection Release”
weeks drive / do a
separate drive test in the
area 495
Drop call failures (SC conflict) C
• Sudden drop to idle Transport
mode (no disconnect channel
messaging) BLER 100%
• Cause of the failure:
overshooting site and
SC reuse
• Short term solution to
add overshooting
neighbour in ADJS
definitions
Cell ABC, SC258
496
Drop Call - Uplink Interference D
UL interference from the
SIB7 message
497
Drop Call – Link Balance D
• UL & DL Power Control commands can UE RX power control
message: DL reception
help indicating problems in link balance. weak -> UE is ordering
WBTS to increase power.
• PC frequency is 1500 Hz, thus ideally
the sum of PC commands to increase or
decrease power is 1500
• E.g. if the sum of UL PC commands is <
1500, this would indicate UE is starting
to loose synchronization
• in Compressed Mode there is less PC
commands, UE spends time on 2G
Sum of UL PC commands
< 1500, UE not receiving all
the PC commands.
498
Drop call failures – System issue RNC or BTS ?
• “CC Disconnect” due to “DL RRC Connection Release” is just a
consequence of failure which can be due to different reasons
– From UE point of view L3-messaging does not identify the point of failure
distinctly
– BTS or RNC failure? => Suspect BTS first, then RNC
• Rule out BTS failures

– Check the site performance from Counters (Iub, Service level, cell resources
SHO, etc) and that site is carrying traffic
– PrxNoise, receive link parameters, alarms
– SC–reuse
– UE performance ?
• Identified causes for Active Set Update failure
– “Deaf” sites (PrxNoise)
– Faulty HW
– SC-reuse
499
Part XIV
HSDPA
500
Content
• HSDPA technology and RAS05.0 features and main parameters
– HSDPA Introduction
– BTS/MAC-hs features
• Mac-hs Protocol
• Physical channel structure
• HSDPA Power
• Link adaptation
• Flow control
• Scheduling
• L1 error correction
– RRM
• Channel type selection
• Power allocation
• HS-DSCH release due to low utilisation or low throughput
• Associated UL DPCH
• Mobility
• HSDPA RAN dimensioning
– HSDPA Dimensioning process
– Air interface capacity dimensioning
– DL Link budget, coverage
– UL link budget, coverage
– WBTS Channel Element dimensioning
– Example dimensioning case
• Planning Process
– Pre-planning with Nokia NetAct planner
– HSDPA design and upgrade options
– Indoor HSDPA solutions
501
Why HSDPA?
Improved
– cell throughput
– maximum user throughput
– round trip time
– spectral efficiency
• Meant for NRT services (background and interactive classes supported first).
• Later releases also RT services (streaming class, VoIP).
502
Faster Content Download Times with HSDPA
100
GPRS 3+2
90 EDGE 3+2
80 WCDMA 384 kbps
70 HSDPA (700 kbps)
WLAN .11b
60
Seconds
50
40
30
20
10
0
30 kB digital 100 kB video 300 kB Symbian 4 MB MP3 Fast
image clip application download of
very large
content
503
HSDPA – General Principle
• Channel quality
information
• Error correction
Ack/Nack •Shared DL data channel
L1 Feedback •Fast link adaptation,
Data scheduling and L-1 error
correction done in BTS
•1-5 codes in RAN05
Terminal 1 (UE) (max.15 codes RAN06)
L1 Feedback
Data •QPSK or 16QAM
modulation
•User may be time
and/or code multiplexed.
Terminal 2 504
Link Adaptation in HSDPA
l
• Adaptive modulation and coding : ig na
in ks
modulation, number of codes and code l k
o wn dbac
rate are optimised dynamically according D e
1) I Fe
to the CQI received from the UE CQ l
2 )
igna
s 3) Adjust modulation
n link scheme according
w
) Do to feedback received
4
from the UE
Good radio conditions use higher order modulation Increased throughput

Q Q
1011 1001 0001 0011

10 00
1010 1000 0000 0010
I
I
11 01 1110 1100 0100 0110
SF16 3840000/16= 240 Ksps 1111 1101 0101 0111 RAN05.0

RAN05.0 QPSK
16QAM GCD1.0
2 bits / symbol
4 bits / symbol
480 kbit/s/HS-PDSCH
960 kbit/s/HS-PDSCH
max. 1.8 Mbit/s (5 codes, 3/4 coding
max. 3.6 Mbit/s (5 codes, 3/4 coding rate)505
rate)
Fast Link Adaptation in HSDPA
16 C/I received C/I varies
Instantaneous EsNo [dB]
14 by UE with fading
12
10
8
6
4
2
0
-2
Modulation
0 20 40 60 80 100 120 140 160
Time [number of TTIs]
16QAM3/4 Link BTS adjusts link adaptation mode
adaptation (Modulation & Coding Scheme)
16QAM2/4 mode with a few ms delay based on
channel quality (CQI) reports
QPSK3/4 from the UE
QPSK2/4
QPSK1/4
506
HSDPA Peak Bit Rates
Coding
Codingrate
rate Coding
Codingrate
rate 55codes
codes 10
10codes
codes 15
15codes
codes
1/4
1/4 600
600kbps
kbps 1.2
1.2Mbps
Mbps 1.8
1.8Mbps
Mbps
QPSK
QPSK 2/4
2/4 1.2
1.2Mbps
Mbps 2.4
2.4Mbps
Mbps 3.6
3.6Mbps
Mbps
3/4
3/4 1.8
1.8Mbps
Mbps 3.6
3.6Mbps
Mbps 5.4
5.4Mbps
Mbps
2/4
2/4 2.4
2.4Mbps
Mbps 4.8
4.8Mbps
Mbps 7.2
7.2Mbps
Mbps
16QAM
16QAM 3/4
3/4 3.6
3.6Mbps
Mbps 7.2
7.2Mbps
Mbps 10.7
10.7Mbps
Mbps
4/4
4/4 4.8
4.8Mbps
Mbps 9.6
9.6Mbps
Mbps 14.4
14.4Mbps
Mbps
240Ksps=480Kbps
480*5*3/4= 1800 Kbps (LAYER 1 THR’PUT !!)
RAN05 (Inter-TTI = 1)
RAN05 GCD1.0 (RN2.1, WBTS WN3.0 ED1) 507

User throughput
• HSDPA uses downlink shared channel.
• Downlink data is scheduled in 2 ms TTIs to users.
• User throughput depends upon the number of active HSDPA users in the
cell/BTS.
2nd TTI
Three HSDPA UEs
receiving data
1st
TTI
3rd TTI
508
Release’99 vs HSDPA Retransmissions
Rel’99 DCH Rel’5 HS-DSCH
RNC
Packet Retransmission RLC Retransmission
RLC
ACK/NACK
BTS
Packet
L1Retransmission
RLC
ACK/NACK L1 ACK/NACK
Terminal
509
New Node B Functionality for HSDPA
RNC Node B Terminals
Packets
Scheduler ARQ &
& Buffer Coding
Flow Control
ACK/NACK &
Feedback
Decoding
New Node B functions:

• Scheduler: Terminal scheduling, Coding & Modulation selection (16QAM as
new modulation)
• ARQ Retransmissions Handling
• Uplink Feedback Decoding
• Flow Control towards SRNC
510
New terminal functionality for HSDPA
RNC Node B Terminal
Packets ARQ Soft Buffer

Decoding & Combining
Flow Control
ACK/NACK &
Feedback
Generation
New terminal functions:

• 16 QAM demodulation
• ARQ Retransmissions Handling
• Soft buffer & combining
• Fast Uplink Feedback Generation & encoding
• First terminals are data cards without voice capability
511
Main HSDPA Features
RAN05 GCD1.0 RAN05.1
Traffic Classes PS interactive PS interactive

Background Background
Modulation QPSK/16QAM QPSK/16QAM
Number of codes for HS-PDSCH 5 codes 5 codes
Scheduler Round Robin Proportional Fair
Mobility Channel type switching to Serving cell change for

DCH on SHO areas associated DCH channels
MultiRAB Not supported AMR+HS-DSCH
Code Multiplexing for HS-PDSCH Not supported Not supported
WBTS processing capability 1 WSPC per BTS 1 WSPC per WCEL (3WSPC
per BTS)
Max number of users per BTS 16 users (QPSK); 12 48 users (16 per cell)
users (16 QAM)
Maximum throughput 3.6 Mbit/s per BTS 3.6 Mbit/s per WCEL
512
Nokia HSDPA Roadmap
513
Nokia HSDPA Roadmap
514
MAC-hs
MAC-hs functions in BTS For each subframe = 3 slots = 2 ms
• Iub Flow Control • Interpret incoming Ack/Nack and CQI
• Data Buffering information.
• Packet Scheduling • Perform Packet Scheduling, i.e. choose
• Link Adaptation one priority queue for transmission.
• Hybrid ARQ – Determine if retransmission is
needed.
• New transmission.
MAC-hs Protocol Services
– Perform Link Adaptation.
• Data transfer on HS-DSCH
– Build MAC-hs PDU.
• In-order delivery
within priority queue • Retransmission
– No need for LA.
– Use same MAC-hs PDU.
• Perform Power Control for HS-SCCH.
SCCH
515
HSDPA Protocol Model
• MAC-d flow RLC RLC

DCH
– UE-specific MAC-d MAC-d
– 1:1 with FP entity DCH DCH
– 1:1 with AAL2 connection DPC FP FP
PHY H PHY
– Up to 8 MAC-d flows per UE
TNL TNL
– RAN05: 1 MAC-d flow per UE
• HS-DSCH UE Uu BTS Iub RNC
– Transport channel
– Shared by the UEs in the cell
– Controlled by MAC-hs RLC RLC
– Between BTS and UE(s) MAC-d MAC-d flow
MAC-d
• HS-PDSCH HS- HS-
HS-DSCH
– Physical Channels MAC-hs MAC-hs DSCH DSCH
FP FP
– 0 to 15 per cell HS-
– RAN05: max. 5 per HSDPA cell PHY PDSCH PHY TNL TNL
• HS-SCCH, HS-DPCCH
UE Uu BTS Iub RNC
– Physical Channles
516
Physical Channels for One HSDPA UE
DL CHANNELS
• HS-PDSCH: High-Speed Physical
BTS Downlink Shared Channel
• HS-SCCH: High-Speed Shared
Control Channel
Rel99 • Associated DPCH, Dedicated
Associated DPCH
Associated DPCH
1-4 x HS-SCCH
DCH Physical Channel.

HS-DPCCH
1-15 x HS-
PDSCH
UL CHANNELS
• Associated DPCH, Dedicated
Physical Channel
• HS-DPCCH: High-Speed Dedicated
Physical Control Channel
UE
517
HSDPA DL Physical Channels
• HS-PDSCH: High-Speed Physical Downlink Shared Channel

– Transfers actual HSDPA data of HS-DSCH transport channel.
– 1-15 code channels.
– QPSK or 16QAM modulation.
– Divided into 2ms TTIs
– Fixed SF16
– Doesn’t have power control
• HS-SCCH: High-Speed Shared Control

Channel Field Number of
uncoded bits
– Includes information to tell the UE how to Channelisation code set information 7 bits
decode the next HS-PDSCH frame
Modulation scheme information 1 bit
– Fixed SF128
Transport block size information 6 bits
– Nokia implementation has power control
Hybrid ARQ process information 3 bits
– Shares downlink power with the HS-
PDSCH Redundancy and constellation version 3 bits
– More than one HS-SCCH required when New data indicator 1 bit
code multiplexing is used UE identity 16 bits
518
HSDPA DL Physical Channels
• Associated DPCH, Dedicated Physical Channel (as many as HSDPA users)
– Transfers signalling (Signalling Radio Bearer (SRB)) information e.g.
RRC measurement control messages
– Fixed SF256
– Power control commands for associated UL DCH
– DPCH needed for each HSDPA UE.
– No SHO support in RAN05 (coming in RAN05.1)
519
HSDPA UL Physical Channels
• HS-DPCCH: High-Speed Dedicated Physical Control Channel
– MAC-hs Ack/Nack information (send when data received).
– Channel Quality Information, CQI reports (send in every 4ms)
– SF 256
• Associated DPCH, Dedicated Physical Channel

– DPCH needed for each HSDPA UE.
– Transfers signalling
– Also transfers uplink data 64, 128, 384kbps, e.g. TCP acks
– No SHO support in RAN05 (coming in RAN05.1)
520
Link Adaptation and Power Control
• The HS-SCCH has power control while the HS-DSCH has link adaptation,
adaptation
e.g. when UE is in poor conditions then HS-SCCH powers-up and the
HS-DSCH bit rate is decreased
– However, overall HSDPA power is ‘fixed’
• Link adaptation and Power control are UE specific
• HS-DSCH link adaptation has typically two loops
– A) Inner loop based on CQI reports from UE
– B) Outer loop in Node-B to control the BLER of the inner loop
• HS-SCCH power control has typically two loops as well
– A) Inner loop based on the fast power control commands from UE
– B) Outer loop in Node-B to control the BLER of the inner loop
– Link adaptation algorithm handles also HS-SCCH power control
• These algorithms are not standardized
HS-DSCH link adaptation
HS-SCCH power control

521
HSDPA TX-Power
• In Nokia implementation:
PtxMaxHSDPA = PHS − PDSCH _ tx + PHS − SCCH _ tx
• Because of Link Adaptation, no actual
power control on HS-PDSCH.
• HS-PDSCH power is set according to
Total HSDPA Power

HS-SCCH power and total HSDPA
power defined by operator (RNC).
• HS-PDSCH power changed slowly to
maintain correct total power over long
averaging window.
• Total power calculated over active
TTIs.
• Because of variations in HS-SCCH HS-SCCH HS-PDSCH
power, instantaneous total HSDPA
power will vary.
• All HS-PDSCH codes (sent to one UE)
must have equal power.
• Interval of HS-PDSCH power
adjustment is defined by HSPDSCH
Power Adjustment Interval (BTS
internal parameter, default 100ms). 522
HS-SCCH Power Control
Inner loop (CQI based) algorithm Outer loop algorithm

• Similar algorithm as for HS-PDSCH • Ack/Nack based.
Link Adaptation, including • Ack and Nack will reduce power.
compensation by outer loop Missing Ack/Nack (DTX) will increase.
algorithm. • Target HS-SCCH BLER is about 1%.
• Using look-up tables, CQI can be
converted to required HS-SCCH
power.
• UE may have different performance
for HS-SCCH and HS-PDSCH. But
outer loop PC should correct this.
• In further releases TPC information
can be utilized too.
Constant HS-SCCH power would cause loss in cell capacity.
523
HS-SCCH Inner Loop Power Control
Node B look-up table
CQI Value λCQI
• In RAN05, the CQI report based HS-SCCH power 1 -5.05
2 -5.68
control shall be used based on the following equation: 3 -6.67
4 -7.73
5 -8.62
PSCCH = PCPICH + Γ + λCQI + P0 6
7
-9.66
-10.5
8 -11.56
where: 9 -12.5
10 -13.57
-Γ is power offset between CPICH and HS-PDSCH. 11 -14.54
Signaled to Node B and UE by RNC. (In Nokia 12 -15.62
13 -16.64
implementation: Γ=PtxMaxHSDPA – PCPICH = PHS- 14 -17.65
15 -18.79
PDSCH+PHS-SCCH – PCPICH) 16 -19.66
17 -20.72
-λCQI is the CQI dependent power offset 18 -21.55
19 -22.83
-P0 is the output of the HS-SCCH outer loop power 20 -23.79
control 21
22
-24.72
-25.65
23 -26.65
24 -27.65
λCQI is defined using a Node B look-up table from CQI 25 -28.65
26 -29.65
27 -30.65
28 -31.65
29 -32.65 524
30 -33.65
HS-SCCH Outer Loop Power Control
• The HS-SCCH outer loop power control is used to fine-tune the power
offset between the HS-PDSCH and HS-SCCH for a given UE.
• When a TB has been transmitted on HS-PDSCH to a UE, the HARQ manager
shall check ACK/NACK feedback from the UE. After the HARQ manager gets
the ACK/NACK feedback or detects the DTX (=no feedback in predefined time)
in uplink HS-DPCCH, it shall inform LA.
• Based on reception of the ACK/NACK the output of the outer loop power control
P0 shall be updated by the following rule:
– If ACK/NACK feedback is received from the UE, implying successful

reception of the HS-SCCH, P0 shall be decreased as:
P0 = max(P0 − Pdown, Pmin )

– If no ACK/NACK is detected, implying unsuccessful reception of the HS-
SCCH by the UE, P0 shall be increased as:
P0 = min(P0 + Pup , Pmax )
– The average BLER on the HS-SCCH (normally 1% target) can be controlled

by selection of the ratio between Pup and Pdown as following:
1 Pmin = -2dB
BLERHS − SCCH = Pmax = 9 dB
Pup, Pdown, Pmin, Pmax are BTS internal parameters Pup
+ 1 Pdown = 0.005 dB
525
Company Confidential Pdown Pup = 0.5 dB
A Two-Step HS-DSCH Link Adaptation (LA)
• Inner loop algorithms

– LA estimates based on UE feedback information, i.e. CQI reports.
– LA estimates based on Node-B measurements = UE power control
commands
– Hybrid schemes using both UE/Node-B measurements.
• Outer loop algorithms

– The goal is to compensate for any bias (random error) introduced by the
inner loop algorithm.
– Bias might be introduced due to offsets in relative UE performance
caused by improved receiver architecture, etc. (CQI calculations
depends on UE)
– The outer loop is based on Ack/Nack.
– The outer loop controls the residual BLER after 1st retransmissions (=2nd
transmission)
– The 1st retransmission is used to take into account the gain from HARQ.
526
A Two-Step HS-DSCH Link Adaptation
RNC Node-B
CQI Outer loop Inner loop HS-DSCH

settings LA algorithm LA algorithm
UL Ack/Nack CQI
detection
527
HS-DSCH Link Adaptation Inner loop
• Step 1: CQI offset compensation

– Compensate for HS-PDSCH transmit power mismatch, if the CQI report
at the UE is derived for a different power level than the one used at the
Node-B.
• Step 2: Mapping of CQI report to feasible LA estimate

– The offset compensated CQI report may not correspond to a feasible LA
estimate, if for instance the number of suggested multi-codes exceeds
the number of HS-PDSCH multi-codes available at the Node-B.
• Step 3: Transport block size adjustment

– Adjust the transport block size for the feasible LA estimate according to
the number of available bits to be transmitted to the UE, i.e., perform a
simple “rounding” operation.
528
HS-DSCH Link Adaptation Outer loop
• The outer loop LA algorithm is used to compensate for the bias introduced in the
inner loop LA algorithm. The output from the outer loop algorithm is the scaling
factor A, which is used in the definition of X for CQI offset compensation
• For each UE, the scaling factor A is set as the initial value Ainit when the UE is
assigned to use HS-DSCH for packet data transmission. After each transmission,
the scaling factor A shall be adjusted based on the ACK/NACK feedback by the
following rules:
– If Ack is received on the first or second transmission, A shall be decreased as:

A = max( A − Astepdown , Amin )
– If NACK is received on the second transmission, A shall be increased as:
A = min(A + Astepup, Amax )
where Astepup, Astepdown, Amin and Amax are the BTS internal parameters
529
CQI Reporting from UE CQI TBS codes M ∆
(dB)
• For CQI reporting the UE measures channel 1 137 1 QPSK 0
2 173 1 QPSK 0
quality of CPICH and assumes that HS- 3 233 1 QPSK 0
PDSCH codes will be transmitted with 4 317 1 QPSK 0
5 377 1 QPSK 0
combined power 6 461 1 QPSK 0
PHS-PDSCH = PCPICH + G + ∆ 7 650 2 QPSK 0
8 792 2 QPSK 0
9 931 2 QPSK 0
10 1262 3 QPSK 0
• Measurement power offset G signalled to 11 1483 3 QPSK 0
Node B and UE by NBAP and RRC (In Nokia 12 1742 3 QPSK 0
13 2279 4 QPSK 0
implementation; Γ=PtxMaxHSDPA - 14 2583 4 QPSK 0
PCPICH) 15 3319 5 QPSK 0
16 3565 5 16QAM 0
• Reference power adjustment D is used to 17 4189 5 16QAM 0
signal how much quality is better than 18 4664 5 16QAM 0
19 5287 5 16QAM 0
maximum TBS supported by UE (given in UE 20 5887 5 16QAM 0
category specific tables). 21 6554 5 16QAM 0
22 7168 5 16QAM 0
• Table to the right (from 25.214) defines CQI 23 7168 5 16QAM -1
for UE categories 1 to 6. Based on 10% 24 7168 5 16QAM -2
25 7168 5 16QAM -3
BLER. 26 7168 5 16QAM -4
• Algorithm to calculate CQI value is not 27 7168 5 16QAM -5
28 7168 5 16QAM -6
specified by 3GPP so CQI reporting can 29 7168 5 16QAM -7
have UE specific differences. 30 7168 5 16QAM 530
-8
CQI Compensations
• The UE reports the CQI measured assuming a transmit power PCPICH + Γ + ∆.
• CQI offset compensation is used to take into account:
– the actual transmit power of HS-PDSCH may be different

– the radio link conditions may have changed after the latest CQI was
measured
• The offset X is used to convert the measured CQI value into a compensated
one using the formula:
– CQIcompensated = CQImeasured + X (X expressed in dB)
where
– X = PHS-PDSCH – (PCPICH + Γ) – A – Y (powers expressed in dB)
where
– A is the output of the Outer Loop LA Algorithm
– Y is the Offset compensation based on TPC. Not supported in WN3.0
(Y=0)
531
CQI Compensations - Example
CQImeasured
X = 3 dB
CQIcompensated
What does it mean in practice?

Higher TBS higher Bitrate
532
Missing CQIs
• If BTS hasn’t receive any CQI report before scheduling data for UE, CQI
value #5 is used to define TFRC (transport format and resource combination).
In this case, the CQI offset compensation will not be applied.
• If MAC-hs detects that several (Nmissed_CQI) consecutive CQI reports from

one UE are not received, MAC-hs stops scheduling of any priority queues of
the UE. The reason for missing CQIs can be HS-DPCCH coverage problem.
• Nmissed_CQI is BTS internal parameter, default 4
533
Link Adaptantion SUMMARY
• HS-DSCH link adaptation algorithm description:
– Nokia’s HS-DSCH link adaptation algorithm is based on an inner loop and
outer loop control mechanism.
– Inner loop algorithm: Adjusts the HS-DSCH transmission format every TTI
based on CQI (Channel quality indicator) measurement reports from the
HSDPA-user terminal.
– Outer loop algorithm: Controls the BLER operating point of the inner loop
algorithm by monitoring the Layer-1 Ack/Nack’s from past transmissions
(and retransmissions).
• Comments on HS-DSCH link adaptation performance:
– The BLER on the first transmission is typically 10%-15%, which results in
a residual BLER of approx. 1% after the first retransmissions when using
HARQ with Chase Combining.
– The probability of having an RLC retransmission triggered after 3 Layer-1
retransmissions is less than 0.5%, and often less than 0.1%.
534
HSDPA Flow Control
BTS Buffers
User data 010101 BTS flow control

from RNC monitors the
110011 BTS buffers and
101010 measures the
101110 throughput for
each queue
011001
001110
101111 S
000011 c
P h
111011 a e
001111 c d
101001 k u
e l
010010 t
000001 e
r
111011
Request for DSCH data, volume dependent on BTS measurements

535
Flow Control
• Algorithms runs on Node B. • RNC chooses when to send the data
Operated per priority queue. during each interval.
• Inputs: • RNC may further limit the amount of
– Amount of data at Node B data to send based on
– Estimated throughput – Node B has allocated more
• Outputs: capacity than there is data to
– Maximum MAC-d PDU size send.
– HS-DSCH credits – Iub capacity is limiting (scheduling
• # packets per interval for this at RNC).
• 0 – 2046 or infinite (2047) • Example
– HS-DSCH interval – credits = 4
• length of each interval – interval = 10 ms
• 0 – 2550 ms, 10 ms
– repetition = 3
resolution
• 10ms default
– HS-DSCH repetition period
10 ms 10 ms 10 ms
• number of intervals
= MAC-d
• 1 – 255 or unlimited PDU 536
• Unlimited (0) default) Company Confidential
Example Simple Flow Control Implementation
• The most simple implementation for the flow controller is to do periodic

capacity allocations. The order of the capacity allocations among the UEs
can be either round robin or based on the MAC-hs or RLC buffer status (or a
combination of both).
• This implementation has some constraints with regards to the guaranteed
bitrate and number of simultaneous users.
UE1 UE1
Scheduling
UE2 UE2
UE3 UE3
RNC buffer Node-B buffer

537
Scheduling
• Fast scheduling : includes more compression to data packets, a faster transmission
time interval (TTI) reduced from 10 ms to 2 ms and the BTS taking care of packet
scheduling
• Round-Robin (RAN05)
– UE in front of queue is scheduled, then moved to back of queue.
– Each UE gets same amount of resources. The throughput depends on link
conditions.
• Proportional Fair Resource (RAN05.1)
– UE with best throughput relative to its mean throughput is chosen.
– On short time-scale, different UEs get different resources.
• Reasons why UE sometimes cannot be
scheduled:
– Some UEs do not support reception in
consecutive TTIs.
– BTS cannot receive Ack/Nack because of
poor uplink quality.
2nd
– Reordering buffer is stalled (too much existing TTI
data pending) 3rd
TTI 1st
– All HARQ processes of UE are full and next TTI538
transmission would be a new one.
Packet Scheduling Strategies
Operator adjustable slopes for
Min. target Throughput

Fair Throughput
different service and user classes
Tradeoff between cell Fair resources
throughput, coverage C/I based
and user fairness
Spectral Efficiency
Throughput Throughput coverage of Throughput coverage of

high C/I based scheduling fair throughput scheduling
low
Low coverage of high data rates 539
Scheduling: Round Robin (FR)
• The round robin scheduling method is used in RAN’05 to select the
transmission turn for each priority queue. The scheduling is made at the cycle
of TTI (2ms) and based on the scheduling counter R of each priority queue.
• The R shall be updated in every TTI for all the priority queues as the following
rule:
– If one priority queue was scheduled in any TTI, R of the priority queue
shall be reset as 0.
– If one priority queue cannot be scheduled in any TTI due to the minimum
inter-TTI interval requirement, R of the priority queue shall remain the same.
– If there is no data in BTS buffer of one priority queue and there is no
pending retransmission in HARQ processes for the priority queue, R of the
priority queue shall remain the same.
– Otherwise, R of the priority queue shall be updated as R=R+1
• At each scheduling interval (2ms), PS shall select the priority queue, which has
the highest R among those priority queues that can be scheduled in next TTI.
• If several priority queues have the same R value, the priority queue, which has
the highest CQI report value, shall be selected.
• The initial value of R for the new priority queue is 0. 540
RAN05.1
PROPORTIONAL FAIR IDEA
The
Thepacket
packetscheduler
schedulerisislocated
located
ininthe
the Node-B. This opensup
Node-B. This opens upfor
for
Channel quality “fast scheduling” optimized
“fast scheduling” optimized
according
accordingtotothethecurrent
currentchannel
channel
(CQI, Ack/Nack, TPC) conditions
conditionswhich
whichcancanyield
yieldlarge
large
gain at cell and user levels.
gain at cell and user levels.
Basic
Basic diversity
diversity mechanism
mechanism
Data utilizedisismulti-user
utilized multi-userdiversity
diversity==
selection
selectiondiversity
diversityamong
amongthe the
active users.
active users.
UE1
TTI 1 TTI 2 TTI 3 TTI 4
Scheduled user
Multi-user selection diversity
(give shared channel to “best” user)
Channel quality
(CQI, Ack/Nack, TPC) Data
UE2
USER 2 Es/N0 USER 1 Es/N0
541
RAN05.1
Proportional Fair Resource (PF)
• Users are scheduled on top of their Metric calculation

fades. • RICQ = TPinst / TPmean
• Time-scale of the fairness is • RICQ = Relative Instantaneous
Channel Quality
controllable by adjusting the
• TPinst = predicted throughput of UE
forgetting factor of the mean for the subframe being scheduled.
throughput averaging. • TPmean = mean throughput of priority
• Benefits over Round-robin queue in the past.
– Cell capacity gain 30-70%. • TPmean is calculated over subframes
when queue can transmit. When
– Higher average throughput queue is not scheduled, throughput
– Smaller HS-SCCH power is zero and TPmean will go down.
needed • Since all priority queues of a UE
have same radio conditions,
– Easier Link Adaptation due to scheduling among different priority
more constant link conditions queues of a single UE reduces in
per UE practice to round-robin (in RAN’06).
• P-FR chosen for RAN05.1.
• RAN05 will use blind scheduling
(like Round-Robin).
542
L1 error correction (HARQ)
• Hybrid ARQ is a combination of
– Forward error correction (channel coding) and
– Automatic Repeat Request (retransmissions).
• HARQ performs retransmissions of MAC-hs PDUs from Node B to UE.
• HARQ processes
– Typically 6 per UE (depends).
– Stop-and-wait ARQ per process.
– Processes operate in parallel.
• NACK feedback status:
– “Yes” means NACK is received for this HARQ process from the UE
– “No” means ACK/NACK has not received yet
– “DTX” means ACK/NACK was not received in predefined time period.
• Transmitter chooses Redundancy Version (RV) for each transmission.
• Receiver performs combining of different transmission of same MAC-hs
PDU.
– Chase Combining.
– Incremental Redundancy.
– Constellation Rearrangement (16QAM only).
543
Fast Retransmission and L1 Error Correction
• Fast retransmission: data retransmissions are now primarily handled in the
BTS in layer 1 (HARQ). TCP layer and RLC layer retransmissions are still
used.
HARQ process
544
HSDPA L1 Retransmissions
• The L1 retransmission procedure (Hybrid ARQ, HARQ) achieves following

– L1 signalling to indicate need for retransmission fast round trip time
facilitated between UE and BTS
– Decoder does not get rid off the received symbols when decoding fails
but combines the new transmission with the old one in the buffer.
• There are two ways of operating:

– A) Identical retransmission (Soft/Chase Combining): where exactly
same bits are transmitted during each transmission for the packet
– B) Non-identical retransmission (Incremental Redundancy): Channel
encoder output is used so that 1st transmission has systematic bits and
less or not parity bits and in case retransmission needed then parity bits
(or more of them) form the second transmission.
545
HSDPA L1 Retransmissions: Chase Combining
Turbo Encoder
Systematic
Parity 1
Parity 2
Rate Matching (Puncturing)
Original transmission Retransmission

Systematic
Parity 1
Parity 2
Chase Combining (at Receiver)
Systematic
Parity 1
Parity 2
546
HSDPA L1 Retransmissions : Incremental Redundancy
Turbo Encoder
Systematic
Parity 1
Parity 2
Rate Matching (Puncturing)
Original transmission Retransmission

Systematic
Parity 1
Parity 2
Incremental Redundancy Combining
Systematic
Parity 1
Parity 2
547
CHANNEL TYPE SELECTION
• Channel type selection flow for PS data
548
CHANNEL TYPE SELECTION
HS-DSCH is selected if ALL of the following conditions are met:
1. Traffic class and traffic handling priority are allowed on HS-DSCH
– The operator can configure which traffic classes and handling priorities are allowed to be used with HSDPA with
HSDSCHQoSclasses (RNC) parameter.
– RAN05 only interactive and background traffic classes are supported
2. UE capability supports HS-DSCH
3. The cell supports HSDPA and HS-DSCH is enabled in the cell
4. No multi-RAB (RAN05) or supported multi-RAB combination (RAN05.1)
– No multiRAB in RAN05
– AMR + HSDPA possible in RAN05.1
5. The number of simultaneous HS-DSCH allocations in the BTS/cell is below the
maximum number.
6. HsdschGuardTimerHO and HsdschGuardTimerLowThroughput guard timers are not
runnig
– Both guard timers are operator-configurable parameters
7. UE is not performing inter-frequency or inter-system measurements
8. Active set size = 1 (RAN05); RAS5.1 if HSDPAMobility enable Active set size can be > 1
9. UE does not have DCHs scheduled with bit rates higher than 0kbps.
10. HS-DSCH physical layer category is supported
11. If there is no existing MAC-d flow in the cell, condition (A or B, depending on the
HSDPApriority parameter) has to be valid.
A) PtxNC<=
PtxNC<=PtxtargetHSDPA
PtxtargetHSDPA (HSDPApriority
(HSDPApriority = 1)
B) Ptxtotal<=
Ptxtotal<=PtxtargetHSDPA
PtxtargetHSDPA (HSDPApriority
(HSDPApriority = 0)
549
Connection setup
• MAC-d flow can be allocated for UE when UE getting capacity request in cell
FACH or in cell DCH 0kbps
• HSDPA tx power (PtxMaxHSDPA) is allocated, when the first MAC-d flow is

entering the cell.
NBAP: Radio Link Procedures

• HSDPA resources can be added with either RL Setup or RL Reconfiguration
and deleted with either RL Deletion or RL Reconfiguration.
550
MAC-d Flow Setup – UE NOT in CELL DCH
UE BTS RNC PS-CN
Connection between UE and CN. UE is in CELL_FACH state, NRT RAB established, NRT RB and SRBs mapped to CCH.
Capacity request UL(MEASUREMENT REPORT) or DL (MAC indicates). HS-DSCH

selected as DL channel type
NBAP:RADIO LINK SETUP (including HS-DSCH,

MAC-d flow & UL DCH )
NBAP:RADIO LINK SETUP RESPONSE
AAL2 Setups for MAC-d flow, SRBs and return

channel
RRC:RADIO BEARER RECONFIGURATION

(addition of HS-DSCH, MAC-d flow and UL DCH of x kbps, state transition to Cell_DCH, RLC reconfiguration, HS-DPCCH & HS-SCCH
configuration)
RRC:RADIO BEARER RECONFIGURATION COMPLETE
UE is in CELL_DCH state. NRT RBs mapped to HS-DSCH in DL and DCH in UL.
551
MAC-d Flow Setup – UE in CELL DCH
552
HSDPA Power Allocation
Range 0…50 dBm
Step 0.1 dB
Default 37.8 dBm (6 W)
Link adaptation
Power control
• PtxMaxHSDPA determines the achievable throughput
• PtxMaxHSDPA = Ptx hs-pdsch + Ptx hs-scch
• PtxMaxHSDPA is sent to BTS in the NBAP Shared Control Channel Procedure at
BTS start
• PtxMaxHSDPA is allocated when 1st Mac-d flow enters the cell
• PtxMaxHSDPA is time shared between HS-DSCH users. If there is already a
Mac-d flow then a new Mac-d flow is always admitted until the limit of 16 users per
BTS
553
• If there are no Mac-d flows PtxMaxHSDPA power is not kept reserved
HSDPA Priority
Node-B Tx power
HSDPAPriority =1
Max power
• Event A: First MAC-d flow entering

the cell
PtxNC<=PtxtargetHSDPA
-> HS-DSCH is selected, otherwise
DCH scheduling. Ptxtotal
• When HSDPA power is in use:
PtxNC target is Ptxtarget. NRT PtxMaxHSDPA Ptx_target
DCH scheduling upto
PtxtargetHSDPA.
• Event B: Ptx_offset_HSDPA
PtxnonHSDPA>=PtxtargetHSDPA PtxnonHSDPA
Ptx_target_HSDPA
+PtxoffsetHSDPA
PtxNC
- Overload control actions for NRT
DCH(s) started
• Event C:
PtxnonHSDPA>=PtxtargetHSDPA
+PtxoffsetHSDPA
- No more NRT DCHs, all MAC-d
flows in the cell are released
A B C 554
HSDPA Priority
HSDPAPriority =2 Node-B Tx power
Max power
• Event A: First MAC-d flow
entering the cell
Ptxtotal<=PtxtargetHSDPA
-> HS-DSCH is selected, otherwise
DCH scheduling.
Ptxtotal
• When HSDPA power is in use:
PtxNC target is Ptxtarget. NRT Ptx_offset
DCH scheduling upto Ptx_target
PtxtargetHSDPA. PtxMaxHSDPA
• Event B:
Ptx_offset_HSDPA
PtxnonHSDPA>=PtxtargetHSDPA
+PtxoffsetHSDPA Ptx_target_HSDPA
- All MAC-d flows in the cell PtxnonHSDPA

released
• Event C: (normal NRT DCH PtxNC
overload control)
Ptxtotal>=Ptxtarget+Ptxoffset
- Overload control actions for
NRT DCH(s) started
A B C 555
PtxtargetHSDPA AND PtxoffsetHSDPA
1) Recommendation could be to keep the ratio between PC headroom and

Variable Power the same after the introduction of HSDPA
2) When HSDPA power is allocated DCH Power increases to keep BLER targets,
so in order to avoid HSDPA power to be allocated and soon released another
recommendation is to find a proper value for PtxoffsetHSDPA:
• DCH power increase << PtxoffsetHSDPA
556
HS-DSCH & MAC-d release
• HS-DSCH (HSDPA Transport Channel) is released when either UTILIZATION or

THROUGHPUT indicates inefficient use of resources.
• Low Utilisation indication is send when
– MAC-d flow throughput is below MACdflowutilRelThr (default 256 bps)
– AND there is no more data in the BTS buffer (normal release).
• Low Throughput indication is send when
– MAC-d flow throughput is below MACdflowthroughputRelThr (default 0 bps)
– Even when there is still data in the RLC buffer but it cannot be sent
(abnormal release).
• If the release of the MAC-d flow is caused by low throughput, the UE is not
allowed to use HS-DSCH until the HS-DSCH guard timer has expired due to low
throughput (RNP parameter HsdschGuardTimerLowThroughput, def. 30s)
• Associated UL DCH needs to be released also (InactivityTimer), when HS-DSCH
is released
557
PDUs in buffer
Throughput
Low utilization indication
MACdflowUtilRelThr
MACdflowthrougputRelThr
MACdflowUtilTimetoTrigger (default 0s)
558
PDUs in buffer
Throughput
Low throughput indication
MACdflowUtilRelThr
MACdflowthrougputRelThr
MACdflowUtilTimetoTrigger MACdflowThroughputTimetoTrigger (default 5s)
559
L3 starts procedure to release UL NRT DCH and MAC-d flow if:
– MAC-d flow has low utilization and UL NRT DCH can be released
Or
– MAC-d flow has low throughput and UL NRT DCH can be released. In
this case UE specific timer HsdschGuardTimerLowThroughput is started.
If MAC-d flow has both low utilization and low throughput conditions valid at the
same time, the functionality described in case of low utilization is followed
HS-DSCH & associated UE moved to Cell FACH

DCHs released
560
Associated UL DPCH bitrate
• Initial bitrate - set by the parameter HSDPAinitialBitrateUL -

corresponds with the initial bitrate for scheduling. Initial bitrate is
allocated as an uplink return channel bitrate when HS-DSCH is selected
as a DL transport channel. However, if HS-DSCH selection is due to UL
capacity request indicating high data amount (TrafVolThresholdULhigh),
the highest possible uplink bitrate is allocated.
• Upgrade of the bit rate of the UL return channel is based on received

capacity requests (traffic volume measurement). Traffic volume
measurement is active when the allocated bitrate is equal to or lower
than initial bitrate.
• Minimum bit rate - set by the parameter HSDPAminAllowedBitrateUL

- determines the lowest allowed bitrate to be allocated when downgrading
DCH bitrate.
561
Associated UL DPCH bitrate – use case 1/3
562
Associated UL DPCH bitrate – use case 2/3
563
Associated uplink DPCH bitrate – use case 3/3
At t1 due to congestion
HSDPA initial bit rate
cannot be allocated so
DCH is scheduled
instead.
In the example also DCH
scheduling is
unsuccessful so at time
t2 another capacity
request is sent.
564
Mobility: Nokia 3G Parameter Database Structure
100
1 RNC
FMCG 32*ADJG
HSDPA
NRT
RT WBTS 100
1
100 48*ADJI HOPG
1
32*ADJS 100 NRT
FMCI RT
WCELL
HSDPA
NRT 100
RT 1
100
HOPI
100
1
1
HOPS NRT
FMCS RT
HSDPA
HSDPA NRT
NRT RT
RT
565
Mobility: FMCx-HOPS-RNC parameter planning
• Events 1F and 6A trigger HS-DSCH release and channel switching to a DCH 0/0
RAB
• Event 1A triggers HS-DSCH release and channel switching to FACH
• Therefore HSDPA HO parameters can be tightened, to maximize HSDPA territory:
– FMCS: AdditionWindow
– FMCS: Addition Time
– FMCS: HHoEcNoThreshold
– FMCS: HHoRscpThreshold
– FMCI: InterfreqUETxPwrThrNrtPS
– FMCG: GSMUETxPwrThrNrtPS
– HOPS: Enable RRC Release = if it is enabled the channel switch to FACH
doesn’t happen when the event 1A is received but when the neighbor cell
CPICH EcNo is above the serving cell CPICH EcNo by negative margins
(HSDPA HOPS parameters ReleaseMarginAverageEcNo and
ReleaseMarginPeakEcNo)
– RNC: SHO HSDPA Capable UE = if it is enabled the SHO for the R5 UEs
when there is only the SRB (DCH0/0) will happen according to the addition
window defined in the HSDPA FMCS set and according to HSDPA HOPS
parameters ReleaseMarginAverageEcNo and ReleaseMarginPeakEcNo 566
Serving Cell Change via cell_FACH RAS05
• Serving Cell Change switches
The conditions preventing HSDPA allocation in the user from HS-DSCH to
HSPDA coverage area to HSDPA capable UE Cell_FACH
are: • Cell Reselection is done in FACH
• More than one serving cell (Soft Handover)
• Multi-RAB combination (not supported)
through Cell Update
• Lack of HSDPA capacity in the cell • HS-DSCH is selected in the new
cell
Cell B
Cell A Switching to UE moves to
the target cell Switching
Service Cell_FACH
by making to HSDPA
in when Cell B is
cellUpdate Cell B
HSDPA “x”dB stronger
HS-DSCH coverage HS-DSCH coverage
Throughput
HSDPA
128kbps or 384kbps according to parameter settings

DCH
64kbps
567
0
Serving Cell Change via Cell-FACH RAS05
• Serving Cell Change switches
the user from HS-DSCH to
Lack of HSDPA capacity in the target
cell
Cell_FACH then to DCCH
with a typical interruption of 3s
Cell A Switching to UE moves to Cell B

Cell_FACH the target cell Switching
Service
when Cell B is by making to R99 Cell
in
“x”dB stronger cellUpdate B
HSDPA
HS-DSCH coverage DCCH coverage
Throughput
Download
complete
HSDPA

DCH
64kbps
0
568
Resumption timer (RAS05+CD1)
The conditions preventing HSDPA allocation in
• HSDPA resumption timer HSPDA coverage area to HSDPA capable UE
switches the user from DCH to are:
• More than one serving cell (Soft Handover)
DSCH when UE enters
HS-DSCH, • Multi-RAB combination (not supported)
HSDPA area • Lack of HSDPA capacity in the cell
Switching to UE moves to Cell B

Cell A
Service Cell_FACH the target cell Switching to
NRT when Cell B is by making HS-DSCH
DCH “x”dB stronger cellUpdate Cell B
R99 coverage HS-DSCH coverage
Throughput
Download
complete
HSDPA

DCH
64kbps (initial bitrate)
0
569
Mobility: Channel type switching
• Event 1A
– If EnableRRCrelease =0 (disable) [HSDPAHopsIdentifier]
• Radio Bearer Reconfiguration to FACH
– If EnableRRCrelease =1 (enable)
• The transition from HS-DSCH to FACH is not triggered directly. Instead,
the UE is instructed to start periodic CPICH Ec/Io measurements. The
RNC parameters ReleaseMarginAverageEcNo (default 2.5dB) and
ReleaseMarginPeakEcNo (default 3.5dB) [HSDPAHopsIdentifier] are
used to trigger the channel type switching to FACH when
– CPICHEcNoneigh>ReleaseMarginAverageEcNo + CPICHEcNoserving or
– CPICHEcNoneigh>ReleaseMarginPeakEcNo + CPICHEcNoserving
• Events 1F and 6A
– Radio Bearer Reconfiguration to DCH 0/0 kbps is allocated but the ISHO/IFHO
measurements are not started immediately. The UE is configured with the
DCH FMCS parameters and measurements are started accordingly.
570
Addition Window (HSDPA FMCS) = 0dB
Addition Time (HSDPA FMCS) = 1280ms
SHO Parameter settings EcNo Filter coefficient (HSDPA FMCS) = 800ms

SHO of the HSDPA Capable UE (RNC) = disabled
Enable RRC release (HSDPA HOPS) = enabled
Release Margin Average EcNo (HSDPA HOPS) = 2dB
Ec/No HSDPA AdditionWindow (0 dB) Release Margin Peak EcNo (HSDPA HOPS) = 3.5 dB
EcNo Averaging Window (HSDPA HOPS) = 8
DCH ReleaseMarginAverageEcNo
AdditionWindow ReleaseMarginPeakEcNo
P CPICH 2
P CPICH 1
DCH AdditionTime HSDPA AdditionTime

time
HSDPA allocation FACH cell update
Measurement Reconfiguration to FACH

Reports (1A) Measurement
Call start Reports (1A)
SHO HSDPA Capable UE 571
disabled Company Confidential
Mobility : RAN05.1 Serving Cell Change RAN05.1
DPCH
• Since MAC-hs in BTS, HS-DSCH does HSDPA
not support Soft Handover. Softer not
supported either.
• But DPCHs of the HSDPA UE can be in

soft/softer handover (multiple RLs).
• One Radio Link of the UE is designated

the ”HS-PDSCH RL”.
572
RAN05.1
Mobility: RAN05.1 Serving Cell Change
• Following reasons can trigger serving cell change:
– Periodical CPICH Ec/No measurement
– Periodical UL SIRerror measurement
• UL SIR error serving cell < ULSIRerrorServCell (-3dB default)
– Measurement event 1B
• Serving HS-DSCH cell is removed from the active set.
– Measurement event 1C
• Serving HS-DSCH cell is replaced with another cell.
– RL failure of the serving HS-DSCH radio link (loss of UL synchronisation)
• Serving HS-DSCH radio link is deleted.
– RL removal due to Rx-Tx time difference
• Serving HS-DSCH radio link is deleted.
573
Mobility: RAN05.1 Serving Cell Change triggered by
CPICH EcNo
CPICH EcNo RAN05.1
cell2
cell3
HS- DSCH Serving Cell Change initiates
after events A and B
HSDPASevCellWindow (2dB)
A. CPICHEcNoservingcell<=CPICHEcNo best cell
HSDPA CPICH EcNo Threshold (-5dB)
– HSDPAServCellWindow
B. CPICHEcNoservingcell<HSDPACPICHEcNothr
eshold
cell1
HSDPA cell change min interval (3s)
A B t
Candidate HS – DSCH Serving Cells are cell 2 and cell 3 as

• CPICH EcNo target cell <= CPICH EcNo best cell – HSDPAServCellWindow
• UL SIR error target cell > HSDPASIRerrortargetcell (-2dB default)
With following priority:
1) The cell, which has already HSDPA power allocated is chosen as the serving HS-DSCH cell.
- If several cells fulfil the criterion, DL CPICH Ec/No determines the order of the cells.
2) The cell, which has the best DL CPICH Ec/No is chosen as the serving HS-DSCH cell.
3) The cell, which has the next best DL CPICH Ec/No is chosen as the serving HS-DSCH cell. 574
Dimensioning process
Typically dimensioning for a RAN RFQ includes following four steps:
1. Traffic engineering
– Operator usually defines a traffic mix with QoS targets
⇒ Estimated simultaneous # connections per bearer/service type
2. BTS configurations
– Number of BTSs is usually given by the operator, but some BTSs can
be added during the dimensioning due to capacity reasons, if the traffic
volumes are very high and the number of carriers/cell is limited.
⇒ Carrier configuration (1+1+1, 2+2+2, 3+3+3 or etc.)
⇒ WSPC configuration (WSPCs for HSDPA, CCH and DCH)
3. Iub configurations
– Number of BTSs, carrier configuration and simultaneous # connections
per bearer needed as an input.
⇒ # E1 per BTS
4. RNC configurations
– Iub results needed as an input.
– Operator usually defines the BTSs or areas that should be covered by
one RNC location.
⇒ # RNC and configurations
575
HSDPA dimensioning overview
Throughput depends on
the signal to interference
ratio.
UL coverage: additional
margin due to CQI and Ack/Nack
Min. throughput at
the cell edge
UL traffic on DPCH
Avg. SINR ->

Iub avg. cell throughput
Node-B
RNC WSPCs (1 or 3) Avg. user throughput ≈
HSDPA power allocation cell throughput / users
RR or PF scheduling
Supported LA (modulation & codes) 576
Air interface capacity dimensioning methods
1. Use simulation results to 2. Use Excel or Netdim Tools.
define BTS configuration
that can support HSDPA
and DCH throughput + More flexibility to estimate
requirements. effect of different traffic
mixes.
+ Possibility to use different G-
+ Very simple approach. factor/little-i distributions
and orthogonality values.
- No flexibility to estimate
effect of different traffic - Requires use of Excel based
mixes. tool or NetDim4.1
- User has to verify that the
- No possibility to see the dimensioning results are in
effect of changing line with the simulation
parameters. results.
Can be used for simple RFQ Can be used for more complex
dimensioning cases, where analysis and for pre-planning.
the main goal is equipment
configuration based pricing.
577
Method 1
Air interface capacity, method 1

• Simplest way to estimate the air interface capacity for dimensioning purposes, is to
use the simulated capacity results directly.
• Simulation results give the avg. cell throughput for following cases
– Shared DCH + HSDPA carrier with 3-9W HSDPA power
• 5 codes
• QPSK and 16QAM
• Cell based scheduling and BTS based scheduling
• RR and PF scheduling
– Dedicated carrier with 10W HSDPA power
• 5 and 10 codes
• QPSK and 16QAM
• Cell based scheduling
• RR and PF scheduling
• Drawbacks of this method:
– DCH only cell capacity is estimated to be 780kbps in the simulations. The
effect of different traffic mixes is difficult to estimate.
578
Air interface – shared carrier (1/2)
Assumptions:
• Macro cell
• Veh-A @ 3kmph
• RR scheduling
• 5codes, QPSK/16QAM at
both Node-B/UE.
7-8 W power for HSDPA • 1 WSPC/cell for HSDPA
provides good
• Cell throughput without
compromise between HSDPA is 780Kbps (DCH
HSDPA bit rates and R99 traffic).
capacity • Using 7-8 W for HSDPA, the
total cell throughput is
increased by a factor
1150/780=1.47
• Note: 16QAM brings only
about 4% gain to avg. cell
throughput compared to
QPSK in macro cell
environment. For this reason
these curves can be used for
both QPSK and 16QAM
cases.
579
Air interface – shared carrier (2/2)
Assumptions:
• Macro cell
• Veh-A @ 3kmph
• PF scheduling
• 5 codes, QPSK/16QAM at
both Node-B/UE.
• 1 WSPC/cell for HSDPA
• Cell throughput without
HSDPA is 780Kbps (DCH
traffic).
• Using 7-8 W for HSDPA,
the total cell throughput is
increased by a factor
1320/780=1.69
• Note: Average HSDPA
traffic must be considerable
(min. 3 simultaneous users)
to achieve gain due to PF
scheduling. If the avg.
number of simultaneous
HSDPA users is 1-2 per cell
then results in previous
slide should be used even
the PF feature would be
enabled in the network.
580
Results with HSDPA-only Traffic
5 HS-PDSCH codes Vehicular-A Pedestrian-A
10W HSDPA Power
Due to using PF Due to using PF
36%
instead of RR instead of RR Gain
Due to 16QAM
Due to 16QAM 27% Gain
14% Gain
4% Gain
HSDPA: HSDPA: HSDPA: HSDPA: HSDPA: HSDPA:

920 kbps 940 kbps 1200 kbps 960 kbps 1100 kbps 1500 kbps
RR RR PF RR RR PF
QPSK only QPSK/16QAM QPSK/16QAM QPSK only QPSK/16QAM QPSK/16QAM
581
Method 2
Air interface capacity, method 2

NetDim4.1
• New features:
• HSDPA dimensioning
• i - distribution
• QoS calculations
• TCP/IP delays, acks and headers
• Header compression
• Complex traffic modeling gives possibility to estimate effect of different services,
end-to-end delays and packet headers.
• Complexity makes the verification of the results more difficult.
• There is no proper user manual available for NetDim4.1.
• Based on HS-DSCH SINR: This is the narrowband signal-to-interference-plus-
noise-ratio after despreading of the HS-PDSCH. Hence, the SINR includes the
SF16 processing gain for the HS-PDSCH and the effect of using orthogonal codes.
• Based on the simulated SINR -> cell throughput mapping.
582
SINR and throughput mapping – Simulation Result
HS-DSCH POWER 7W (OF 15W), 5 CODES,
• The single-user 1RX-1TX, 6MS/1DB LA DELAY/ERROR
4.0
HSDPA throughput
Average single-user throughput [Mbps]

Rake, Ped-A, 3km/h ACTIVITY FACTOR 100%
versus its average HS- 3.5
DSCH SINR is plotted. Rake, Veh-A, 3km/h
• Notice that these 3.0 Rake, Ped-B, 3km/h
results include the MMSE, Ped-A, 3km/h
2.5 MMSE, Ped-B, 3km/h
effect of fast fading
and dynamic HS- Rake, Veh-A, 30km/h
2.0
DSCH link adaptation.
• An average HS-DSCH 1.5
SINR of 23 dB is 1.0
required to achieve the FROM
maximum data rate of 0.5 WALLU
3.6 Mbps with 5 HS-
PDSCH codes. 0
-10 -5 0 5 10 15 20 25 30
Average HS-DSCH
Average SINR SINR [dB]
(1 HS-PDSCH) [dB]
583
HSDPA DL LINK BUDGET DL DCH LINK BUDGET
• The estimation of HSDPA PathLoss (coverage)

can be obtained from HSDPA link budget:
(below example, find the HSDPA max TX power
in order to get similar DL DCH384 coverage)
Service HSDPA384
Transmitter - Node B
Max Tx Power (HSDPA) 12 W
Max Tx Power (HSDPA) 40.8 dBm
Tx Antenna Gain 18 dBi
Cable Loss 3 dB
EIRP 55.8 dBm
Receiver - Handset
Handset Noise Figure 8 dB
Thermal Noise -108 dBm
Background RSSI -100 dBm
Downlink Load 80 %
Interference Margin 7.0 dB
Interference Floor -93.0 dBm
Min SINR 4.5 dB PHS−PDSCH_ tx = PtxMaxHSDPA − PHS−SCCH_ tx
Service PG 12.0 dB
Rx Antenna Gain 0 dBi
Body Loss 0 dB SINR TO MATCH THE EDGE BITRATE
Receiver Sensitivity -100.6 dB
DL Fast Fade Margin 0 dB
DL Soft Handover Gain 0 dB Service PG = 10 ∗ log10 (16) ; SF = 16
Max. Path Loss 156.3 dB
C = SINR− SF16 + I = SINR− SF16 + InterferenceFloor

NO SHO for HS-DSCH
584
HSDPA DL LINK BUDGET DL DCH LINK BUDGET
• Absolute minimum requirement for SINR

is 0 dB (≈100kbps throughput)
Service HSDPA384
Transmitter - Node B
Max Tx Power (HSDPA) 12 W
Max Tx Power (HSDPA) 40.8 dBm
Tx Antenna Gain 18 dBi
Cable Loss 3 dB
EIRP 55.8 dBm
Receiver - Handset
Handset Noise Figure 8 dB
Thermal Noise -108 dBm
Background RSSI -100 dBm
Downlink Load 80 %
Interference Margin 7.0 dB
Interference Floor -93.0 dBm
Min SINR 0 dB PHS−PDSCH_ tx = PtxMaxHSDPA − PHS−SCCH_ tx
Service PG 12.0 dB
Rx Antenna Gain 0 dBi
Body Loss 0 dB
Receiver Sensitivity -105.1 dB Minimum requirement for SINR
DL Fast Fade Margin 0 dB
DL Soft Handover Gain 0 dB
Max. Path Loss dB
160.8
C = SINR− SF16 + I = SINR− SF16 + InterferenceFloor
NO SHO for HS-DSCH

585
Uplink HS-DPCCH Power Control
• The HS-DPCCH is power controlled relative to the uplink DPCCH every slot
period.
HS-DPCCH Ack/Nack CQI report
∆ACK; ∆NACK ∆CQI ∆CQI
DPCCH
⎛ ⎛⎜ ∆HS−20DPCCH ⎞⎟ ⎞ ∆
∆ HS −DPCCH = 6 10⋅ log10 ⎜10⎝ ⎠⎟
= HS−DPCCH = 10⋅ log10 (30 / 15) = 3
⎜ ⎟ 2
⎝ ⎠
• The power offset parameters [DACK; DNACK; DCQI] are controlled by the RNC and
reported to the UE using higher layer signalling. The power offset can be up to 6
dB within 9 possible quantization steps, see the tables in 25.213.
• The power offsets [DACK; DNACK; DCQI] are radio network planning parameters that
should be selected to fulfill the required UL HS-PDCCH detection probabilities.
586
UL link budget: HS-DPCCH overhead Summary
So, for the available UL return channel bit rates, we derive:

• UL 64
– Delta_ack_dB = 2
– r2 (dB) = -5.46 Offset (dB) = 1.3
• UL 128
– r2 (dB) = -9.54 Offset (dB) = 0.6
• UL384
– r2 (dB) = -9.54 Offset (dB) = 0.4
587
HSDPA BTS Configuration in RAN05
• WSPC with HSDPA SW card required
• All BTS models/configurations, that can
accommodate at least one WSPC,
support HSDPA.
4 users
• One WSPC unit supports max. 3 HSDPA
cells.
• One WSPC unit supports max. 16
8 users 4 users HSDPA users. 1)
• Those 16 HSDPA users can be divided
freely between three cells.
• Packet Scheduler in MAC-hs schedules
all users equally regardless of their cell.
-> Cells get resources depending on the
amount of users and traffic in each.
• Max. one HSDPA WSPC in BTS
-> Max. three HSDPA cells in BTS.
Note 1: The actual amount of supported HSDPA users per
WSPC depends heavily on used Uplink DCH channel speed
associated with HSDPA user, see dimensioning rules
588
HSDPA activation : Channel Allocation in WSPC
WSP type CEs
• Channel Elements (CEs) are used as a
medium to explain resources (processing WSPA 32
power) on WSP cards and Node B.
• Only WSPC supports HSDPA WSPC 64
• WSPC card supports 64 CEs

WSPD 32
• 16 CEs are needed for R99 common
channels for 1+1+1 BTS configuration WSPE 16
• CEs for dedicated channels depend on bit
rate:
User data Rate CE
16 kbps or AMR 1 CE
voice
64 kbps 4 CEs
128 kbps 4 CEs
384 kbps 16 CEs
589
• RAN05: Max. one HSDPA WSPC in BTS -> Max. three HSDPA cells in BTS.
• RAN05.1: max. 3 HSDPA WSPCs in BTS (one WSPC per sector), max 48
HSDPA users/BTS
• 32 CEs are needed to handle HSDPA shared channels. These CEs are
reserved by RRM when HSDPA is enabled at cell start up. One to three
HSDPA cells are supported on one WSPC, capacity is shared between cells
within the WSPC.
1. WSPC 2. WSPC Nth WSPC
Common channels
Carrier
Carrier11 Common chs HSDPA BLOCK
(5codes)
Reserved
Carrier
Carrier11 remaining by RRM
…. capacity
capacity 64 CE
48 CE remaining
capacity
Carrier
Carrier11 32 CE
= R5 HSDPA
Traffic channels
= R99 DCH
590
• The CE reserved in the WSPC for
HSDPA are 32
WSPC 1 WSPC 2
• The WSPC for the associated DCH is DL UL DL UL
selected as for any DCH (Node B sees 1 1
2 2
no difference). 3 3
4 4
• MAC-d flow and DCH of the same UE 5 8 9 10 11 9 9 9 9 5
can be under different WSPC and/or 6 12 13 14 15 10 10 10 10 6
WAM. 7 16 11 11 11 11 7
8 1 1 1 1 12 12 12 12 8
9 1 1 1 1 16 16 16 16 9 1 2 3 4 1 1 1 1
10 1 1 1 1 14 14 14 14 10 5 6 7 2 2 2 2
11 1 1 1 1 15 15 15 15 11 3 3 3 3
12 2 2 2 2 16 16 16 16 12 4 4 4 4
Associated dedicated channels are 13 2 2 2 2 1 1 1 1 13 5 5 5 5
required for HSDPA users: 16 14 2 2 2 2 2 2 2 2 14 6 6 6 6
15 2 2 2 2 15 7 7 7 7
kbps DL for uplink power control 16 16 8 8 8 8
and 64…384 kbps UL channel for
Common channels (for 3 carriers)
TCP ACK’s and uplink user data.
HSDPA 5 codes
HSDPA and associated DCHs HSDPA associated DCH

effectively balance the base band (16 users, 16 kbps DL, 128 kbps UL)
UL/DL resource consumption even 384 / 64 kbps DCH users (2 users)

if the traffic is DL oriented.
591
WSPC capacity calculations- check
WSPC capacity:
• HSDPA with 5 HS-PDSCH codes in RAN05:
• 16 HSDPA users (16 uplink HS-DPCCHs and 3.6 Mbit HSDPA
downlink)
• 32 DCHs at 16 kbit/s OR 16 DCHs + common channels
• Use following method to calculate needed processing capacity:

• Calculate DL CEs = Rel99 traffic DL CEs + associated DL DCH CEs
• Calculate UL CEs = Rel99 traffic UL CEs + associated UL DCH CEs
• Select bigger one of DL CEs / UL CEs and add CEs for CCHs
• Add required HSDPA CEs (32)
– Associated DL DCH (3.4 kbit/s SRB) takes capacity of one CE per

HSDPA user.
– Associated UL DCH takes capacity related to its data rate: 4/4/16
CEs per user for data rates 64/128/384kbps.
592
WSPC capacity, example
INPUTS:
If the Rel99 traffic requires 30 CE in DL and 20 in UL (SHO OH included).
The average number of HSDPA users is assumed to be 5 and associated UL
traffic is carried on 64kbps bearer.
One WSPC card is allocated for HSDPA. BTS configuration is 1+1+1.
RESULT:
• DL CEs = 30 +5 and UL CEs= 20+4*5=40
• Total CEs for traffic + CCHs = 40 +16 =56
• HSDPA takes 32 CEs, so totally we need capacity of 56+32=88 CEs.
• For this capacity two WSPC cards is required (or WSPC + WSPA/D).
593
Pre-planning with NetAct Planner
• NetAct Planner Version 5.1 has HSDPA support
• Previous version of NetAct Planner (V4.2) can be used for simple HSDPA
coverage planning by using Pilot Ec/Io –> SINR -> HSDPA throughput
mapping tables.
594
Pre-planning with NetAct Planner V4.2
• Static Pilot Ec/Io can be mapped to SINR when we know:
– Pilot power
– HSDPA power
– Total BTS power
– DL orthogonality
• SINR vs. throughput lookup table gives the HSDPA throughput
• Draw back is that requires that all BTSs in the analysis area have similar
power assumptions/parameters.
SFHS-PDSCH= spreading factor (16)

PHSDPA= HSDPA power
PHSDPA
ρHSDSCH = SF16 Ppilot= Pilot Power
Ppilot Ptot= Total BTS tx power
− αPtot
ρ pilot α= DL orthogonality
ρpilot = Pilot Ec/Io
ρHSDSCH = HSDPA SINR
595
596
• Example for 5 codes with 16QAM:
– BTS total power 14W
– Pilot power 2W
4 W HSDPA power 6 W HSDPA power
Pilot HSDPA HSDPA, 5codes HSDPA HSDPA, 5codes
Ec/Io SINR Throughput SINR Throughput
-15 0.56 167 2.32 249
-14.5 1.13 193 2.89 279
-14 1.70 220 3.46 316
-13.5 2.29 248 4.05 354
-13 2.89 279 4.65 393
-12.5 3.50 319 5.26 432
-12 4.14 359 5.90 479
-11.5 4.79 401 6.55 540
-11 5.47 445 7.23 604
-10.5 6.18 506 7.94 671
-10 6.92 576 8.68 745
-9.5 7.72 650 9.48 865
-9 8.57 731 10.34 995
-8.5 9.51 871 11.27 1137
-8 10.57 1030 12.33 1278
-7.5 11.78 1208 13.54 1437
-7 13.26 1400 15.02 1688
-6.5 15.20 1725 16.96 2092
-6 18.23 2415 19.99 2871
-5.5 28.23 3619 29.99 3619 597
HSDPA planning process
1. Monitoring and analyzing existing R99 network Monitoring and analyzing
performance existing R’99 network
performance
• Average used BTS power
• Average BTS WSP card utilization
• SHO overhead
2. Dimensioning, pre planning and deployment strategy HSDPA performance targets
– BTS power resources for HSDPA and deployment strategy
– HSDPA strategy: hot spot or whole area coverage
– Dedicated or shared carrier
3. Parameter planning
Detailed BTS andIub
– Priority parameter configurations and RNC
– Power thresholds planning
– Shared carrier
• Power allocation for HSDPA
– Dedicated carrier
HSDPA parameter planning
• Directed RRC in cell/cluster level
– Mobility
• SHO parameters
4. Performance monitoring
– Counters
Performance monitoring
– Online monitoring
598
HSDPA Deployment Strategy - Method 1
Given DCH traffic
requirements assigns
to HSDPA the power
that is not used for
DCH and obtains the
HSDPA cell
throughput and
WBTS configuration
(number of WSPC,
number of carriers)
599
HSDPA Deployment Strategy - Method 2
Given HSDPA cell

throughput derives RAS05.1
the corresponding
HSDPA
transmission power
and WBTS
configuration
(number of WSPC,
number of carriers)
600
Site/cell design possibilities
• Hotspot configuration
– HSDPA is set up in an isolated cell where HSDPA users are more likely to
be present
– In most cases a WSPC can be allocated to handle the (unique) HSDPA cell.
– Dual carrier configuration can be enabled, the second carrier mostly
dedicated to HSDPA users
• General HSDPA overlay

– Shared carrier
• In low loaded/pre-launch networks
• If HSDPA does not have priority over NRT DCH and is used as
complement
• Expectation of low initial HSDPA traffic
• In case of low/medium HSDPA traffic requirements by operator
– Dedicated HSDPA carrier on f2, most DCH users on f1
• In loaded networks
• HSDPA seen as priority, high desired HSDPA throughput
601
General HSDPA deployment
Phase 1: Low DCH and HSDPA traffic
– A single carrier shared between DCH and HSDPA.
– Nokia RAN´05 HSDPA solution allows to choose whether prioritize Non-Real-Time DCH users over HSDPA users. In
any case Real-Time CS users (Voice, Video) are prioritized over PS Non-Real-Time users.
– Exact resource (BTS Output Power,Channel elements, RNC resources) allocation between HSDPA and DCH users
can be derived from HSDPA throughput requirements.
– Additionally HSDPA hotspots can be created (Indoor, Microcells) to handle specific, high traffic areas (Company
buildings).
Phase 2: Medium to high RT DCH traffic, medium NRT DCH and HSDPA traffic.
– Second carrier added on high traffic sites.
– Based on traffic predictions and network KPIs measurements, additional carriers can be set up on specific sites or at
cluster level.
– In case of multiple carriers, HSDPA users can be directed to a dedicated or semi dedicated carrier.
buildings).
Phase 2: High RT DCH traffic, NRT DCH and HSDPA traffic.

– Two to three carriers set up on whole high traffic areas.
– HSDPA users can be directed to a dedicated or semi dedicated carrier.
buildings).
602
Scenario 1: Carrier shared between HSDPA and R99
= R5 HSDPA • Resource sharing between HSDPA and R99

= R99 DCH • Most likely initial network setup when low DCH
and HSDPA traffic
☺cost efficiency
f1
f1 ☺Simple parameter planning (traffic handling)
Capacity (additional load from HSDPA)
Impact exist Rel 99 performance (noise rise)
Main Planning parameters:

• HSDPA priority over NRT DCH
• PtxHSDPAmax : the higher the power for HSDPA, the higher the cell
throughput but also the higher the drop in Ec/N0 for R99 users. Power left
from R99 DCH should be estimated by power measurement statistics
reported in the cell resource table
603
Scenario 2 : Dedicated carrier to R99 UEs + Shared
carrier between R99 and HSDPA UEs
• UE moved by RRC connection setup according to
UE release R5 in RRC Connection Setup
message
• Recommended for hotspot, indoor, major
customer office and also in mature network
☺Absorb more traffic for rel99
f2 ☺Less impact to R99 than first solution
f2 = R5 HSDPA+R5 DCH
☺More capacity and more cell throughput
f1
f1
= R99 DCH ☺Additional WSPC card could be required, but
this will avoid the requirement for a
subsequent visit for a subsequent capacity
Main Planning parameters: upgrade
Cost
• HSDPA priority over NRT
DCH More WSP&Iub&DMPG? overhead from 2nd
carrier activation
• PtxHSDPAmax
Additional HW required
• Directed RRC Setup for R5
More complex parameter planning (DRRC,
UE capability, traffic balancing)
604
Directed RRC setup for HSDPA layer
• All idle mobiles are forced to camp on f1
• Any UE reporting Rel’5 is directed to HSDPA layer, others to Rel’99 layer
Rel’5
RRC setup on f2
f2, HSDPA
All idle
R’99
UEs camp
RRC setup onto f1
on f1
f1, Rel’99
605
DirectedRRCEnabled
Enables / disables the use of Directed RRC connection setup or Directed RRC
connection setup for HSDPA layer in the cell.
0: Directed RRC connection setup is not enabled in the cell

1: Directed RRC connection setup is enabled in the cell
2: Directed RRC connection setup for HSDPA layer is enabled in the cell
Description of parameter relationships:

If this parameter value is 1 then Directed RRC Connection Setup from a loaded cell
to a cell belonging to the same sector but on different frequency that is less
loaded is activated (Release 4 feature).
If this parameter value is 2 then Directed RRC Connection Setup from a cell that
doesn’t support HSDPA to a cell belonging to the same sector but on different
frequency that supports HSDPA is activated for all UE accessing the service that
are Release 5 (Access Stratum Release Indicator = R-5 in the RRC Connection
Request Message).
606
Conditions to activate the feature :
• There must not be more than two layers. This is checked from SectorID (no
more than 2 same SectorIDs per BTS exist in the cells where
DirectedRRCEnabled=2).
• Following WCEL parameters have to be the same in the two cells:

– Transmission power of the primary CPICH channel (PtxPrimaryCPICH)
– Target for transmitted power (PtxTarget)
– Offset for transmitted power (PtxOffset)
– Target for received power (PrxTarget)
– Offset for received power (PrxOffset)
– Mobile Country Code (MCC)
– Mobile Network Code (MNC)
607
HSDPA within the Indoor Environment
• Indoor solutions are able to offer potentially high average HSDPA throughput
• Throughput performance depends upon SINR conditions as for macrocells
• SINR conditions can be good if
– there is good isolation from the macrocell layer
– a dedicated RF carrier is used
– high power is assigned to the HS-PDSCH
• The indoor solution may use

either an active or passive DAS
Omni Directiona
• Active DAS should be selected if l
passive DAS losses are relatively
high, i.e. 18 dBm cannot be
maintained at each remote
Direction
antenna connector
al
608
HSDPA within the Indoor Environment
Differences between DAS and Nokia AIR Solution
• DAS - Currently most commonly used indoor solution is coaxial cable based
Distributed Antenna System. Any Nokia BTS can be used to provide DAS
solution.
• AIR - Nokia Advanced Indoor Radio - is fibre optical based active distributed
antenna system, which can be integrated within Nokia Ultrasite WCDMA BTS.
DAS AIR
Cable distribution Coaxial cable based Optical cable based
Supported, but not recommended- need to
Diversity duplicate all DAS component (expensive) Supported
System Sharing Can be used to share different technologies Support only WCDMA
All DAS components, except BTS, are passive
Maintenance & Power so no maintenance are required Active components
Base Station Any Nokia BTS can be used with DAS Currently only Ultrasite
HSDPA Supported Supported
609
RF Carrier Assignment for Indoor Solutions
• The design of an indoor solution should start with measurement of the existing
macrocell coverage
• Strong macrocell coverage will degrade the indoor solution SINR
• If macrocell coverage is strong then there may be a requirement to assign a
dedicated RF carrier to the indoor solution
Benefits Drawbacks
RF Carrier shared Soft handover may be used Soft handover may trigger channel type
with macrocells to avoid hard handovers switching from HS-PDSCH to DPCH
SINR is likely to be Hard handover required
Dedicated RF Carrier
greater
• The indoor solution can be assigned 2 RF carriers: 1st RF carrier for DPCH
connections and 2nd RF carrier for HSDPA connections
• In the case of active DAS, assigning 2 RF carriers will reduce the maximum
downlink transmit power per carrier
610
Indoor Solution Link Budget
• The indoor solution HSDPA link budget should account for the
– low downlink transmit power radiated by the DAS
– high orthogonality reducing the level of own cell interference
– lack of soft handover gain
– no fast fade margin due to no power control
– reduced slow fade margin
– potentially high level of intercell interference from the macrocell layer
• The HSDPA link budget should be completed in parallel to the DPCH link
budget
• The average HSDPA throughput should be evaluated based upon link budget
results and the level of intercell interference from the macrocell layer
• It may be necessary to increase the number of indoor solution antennas to
improve the HSDPA SINR conditions
PHSDPA
• SINR conditions should SF16 ×
be evaluated for a range C SINR
SINR = PathLoss
of distances from the =
I SF16 ⎛ PBS ⎞
indoor solution antennas ⎜ ⎟ × α + I Other + PNoise
⎝ PathLoss⎠
611
HSDPA Parameters for the Indoor Environment
• The HSDPA parameter set for the indoor environment can be the same as that
for the macrocell environment
• The transmit power assigned to HSDPA must be sufficient to achieve adequate
SINR and downlink throughput
• The HSDPA priority should be configured to maintain the HSDPA connections
rather than the DPCH connections
• FMCS, FMCI, FMCG and HOPS parameter sets should be defined

specifically for HSDPA RNC
• The soft handover parameters and the inter-frequency and inter- WBTS
system handover parameters should ensure that channel switching
to DPCH does not occur unnecessarily WCEL
ADJS
FMCS
FMCI
FMCG
HOPS
612
3GPP TSG RAN references
• WG1 specifications
TS 25.201 Physical layer - General description (Release 5)
TS 25.211 Physical channels and mapping of transport channels onto physical channels (Release 5)
TS 25.212 Multiplexing and channel coding (Release 5)
TS 25.213 Spreading and modulation (Release 5)
TS 25.214 Physical layer procedures (Release 5)
TS 25.306 UE Radio Access capabilities (Release 5)
TS 25.308 High Speed Downlink Packet Access (HSDPA) Overall description; stage 2 (Release 5)
TS 25.321 Medium Access Control (MAC) protocol specification (Release 5)
TS 25.331 Radio Resource Control (RRC); protocol specification (Release 5)
TS 25.423 UTRAN Iur Interface RNSAP Signalling (Release 5)
TS25.425 UTRAN Iur Interface user plane protocols (Release 5)
TS 25.433 UTRAN Iub Interface NBAP Signalling (Release 5)
TS 25.435 UTRAN Iub Interface User Plane Protocols for Common Transport Channel Data Streams
(Release 5)
TS 25.101 UE Radio Transmission and Reception (Release 5)
• Technical Reports
TR 25.848 Physical layer aspects of UTRA High Speed Downlink Packet Access (Release 4)
TR 25.855 High Speed Downlink Packet Access; Overall UTRAN Description (Release 5)
TR 25.950 UTRA High Speed Downlink Packet Access (Release 4)
613

KL Workshop PDF

Hochgeladen von

Dokumentinformationen

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

KL Workshop PDF

Hochgeladen von

Copyright:

Verfügbare Formate

Part I

Multiservice Environment Air Interface

BTS BTS Node B Node B

Microsoft Word SIM-ME i/f or Cu

TDD FDD MSS TDD FDD MSS

FDD Mode TDD Mode 5

•Power control: transmissions by the UE must be carefully controlled so that all

• TD-SCDMA Technical Summary

Conversational Streaming Interactive Background

• Video phone (WCDMA)

• Interactive traffic - fundamental characteristics for QoS:

• Background traffic - fundamental characteristics for QoS:

WCDMA TDMA (GSM)

Correlation Narrowband Interference Spread the power

Voice user (12,2 kbit/s) (User data rate) x (spreading ratio)=

Frequency (Hz) •Spreading sequences of

Data Channel coding 3.84 Mcps 3.84 Mcps

Channelisation code (OVSF)

12.2 Kbps Uplink Reference channel

64 Kbps Uplink Reference channel

Transmitted Received signal Time and phase Combined

• High level of multipath diversity

Rake Finger RESOLUTION = 0.26µs 78m

Logical Channels (define what type of data is transferred ) RLC

Physical Channels Frames PHY

Nt DC L3 RRC (Radio Resource Control): Broadcast of information

Bearers L2/PDCP (Packet Data Convergence Protocol): Header

L2/BMC (Broadcast/Multicast Control): Storage of Cell broadcast

L2/RLC (Radio Link Control): Segmentation (Reassembly),

Logical L2/MAC (Medium Access Control): Mapping Logical channels

Common Transport Channels:

Dedicated Transport Channels:

BCCH BCH P-CCPCH

PCCH PCH S-CCPCH

CCCH RACH PRACH

• Transport channels contain the data generated at the higher layers,

TTI TTI TTI

TB Transport Block TF Transport Format

•Primary Synchronisation Channel (P-SCH) (SLOT and CHIP SYNCHRONIZATION)

2560 Chips 256 Chips

Secondary Synchronisation Channel (S-SCH)

Cs1 Cs2 Cs15 Cs1

Slot 0 Slot 1 Slot 14 Slot 0

CPICH Ec/No = CPICH RSCP

CPICH Ec/No CPICH RSCP UTRA carrier RSSI

Ec/No values in dB RSCP values in dBm RSSI values in dBm

Finally, I get the

In GSM all common channels have the same power,

The S-CCPCH can be used to transmit the transport channels

• Forward Access Channel (FACH) and

• The Secondary CCPCH (Common Control Physical Channel) carries

• DL common Channel configuration in case of three SCCPCH For SAB

Logical channel DTCH DCCH CCCH BCCH CTCH PCCH

Transport FACH-u FACH-c FACH-c FACH-s PCH

Physical channel SCCPCH SCCPCH SCCPCH

• Recommended value of the transmission power of the SCCPCH is depended on the

Slot 0 Slot 1 Slot 2 Slot 14

• carries PCH and FACH

0 S-CCPCH carrying one PCH

1 S-CCPCH carrying one PCH

K-1 S-CCPCH carrying one PCH

S-CCPCH without PCH

for paging indication no transmission