ADVANTAGES OF CELLULAR COMMUNICATIONS

Cellular networks have many advantages over the existing land telephone networks.
There are advantages for the network provider as well as the mobile subscriber.
Advantages to the customer

Mobility
Flexibility
Convenience
Advantages to service provider
etwork expansion flexibility
!rofit Margin
"fficiency
"asier #econfiguration

CELL SIZE
The number of cells in any geographic area is determined by the number of M$
subscribers who will be operating in that area% and the geographic layout of the area
&hills% lakes% buildings etc'.
LARGE CELLS
The maximum
cell si(e for
)$M is
approximately
*+ km in
diameter% but
this is dependent on the terrain the cell is covering and the power class of the M$. ,n
)$M% the
M$ can be transmitting anything up to - .atts/ obviously% the higher the power output of
the M$ the larger the cell si(e. ,f the cell site is on top of a hill% with no obstructions for
miles% then the radio waves will travel much further than if the cell site was in the middle
of a city% with many high0rise buildings blocking the path of the radio waves.
)enerally large cells are employed in1
#emote areas.
Coastal regions
2reas with few subscribers
3arge areas which need to be covered with the minimum number of cell sites.
SMALL CELLS
$mall cells are used where there is a re4uirement to support
a large number of M$s% in a small geographic region% or
where a low transmission power may be re4uired to reduce
the effects of interference. $mall cells currently cover 5++ m
and upwards.
Typical uses of small cells1
6rban areas.
3ow transmission power re4uired
7igh number of M$s.
.
THE TRADE OFF -LARGE VS SMALL
There is no right answer when choosing the type of cell to use. etwork providers would
like to use large cells to reduce installation and maintenance cost% but reali(e that to
provide a 4uality service to their customers% they have to consider many factors% such as
terrain% transmission power re4uired% number of M$s etc. This inevitably leads to a
mixture of both large and small cells.

SECTORIZATION
The cells we have looked at up to now are called omni0directional cells. That is each site
has a single cell and that cell has a single transmit antenna which radiates the radio
waves to 89+ degrees.
The problem with employing omni0directional cells is that as the number of M$s
increases in the same geographical region% we have to increase the number of cells to
meet the demand. To do this% as we have seen% we have to decrease the si(e of the cell
and fit more cells into this geographical area. 6sing omni0directional cells we can only
go
so far before we start introducing co0channel and ad:acent channel interference% both of
which degrade the cellular network;s performance.
To gain a further increase in capacity within the geographic area we can employ a
techni4ue called sectori(ation. $ectori(ation splits a single site into a number of cells%
each cell has transmit and receive antennas and behaves as an independent cell.
"ach cell uses special directional antennas to ensure that the radio propagation from one
cell is concentrated in a particular direction. This has a number of advantages1 firstly% as
we are now concentrating all the energy from the cell in a smaller area 9+% <5+% <-+
degrees instead of 89+ degrees% we get a much stronger signal% which is beneficial in
locations such as in0building coverage. $econdly% we can now use the same
fre4uencies in a much closer re0use pattern% thus allowing more cells in our geographic
region which allows us to support more M$s.

USING SECTORED SITES
The distribution of #F carriers% and the si(e of the cells% is selected to achieve a balance
between avoiding co0channel interference by geographically separating cells using the
same #F fre4uencies% and achieving a channel density sufficient to satisfy the anticipated
demand.
The diagram opposite illustrates how% by sectoring a site we can fit more cells into the
same geographical area% thus increasing the number of M$ subscribers who can gain
access and use the cellular network.
This sectori(ation of sites typically occurs in densely populated areas% or where a high
demand of M$s is anticipated% such as conference centres=business premises.
4 SITE/3 CELL
2 typical re0use pattern used in )$M planning is the > site=8 cell.
For example% the network provider has 89 fre4uencies available% and wishes to use the >
site=8 cell re0use pattern he may split the fre4uencies up as follows1
,n this configuration each cell has a total of 8 carriers and each site has a total of ?
carriers. ,f the provider wished to reconfigure to a 8 site=8 cell then the result would be1
2s can be seen from the table% each cell now has > carriers and each site has <5
carriers . This has the benefit of supporting more subscribers in the same geographic
region % but problems could arise with co0channel and ad:acent channel interference.
4 site/3 cell
FREQUENCY SPECTRUM
The fre4uency spectrum is very congested% with only narrow slots of bandwidth allocated
for cellular communications. The list opposite shows the number of fre4uencies and
spectrum allocated for )$M% "xtended )$M ?++ &")$M'% )$M <-++ &@C$<-++' and
!C$<?++.
2 single 2bsolute #adio Fre4uency Channel umber &2#FC' or #F carrier is actually
a pair of fre4uencies% one used in each direction &transmit and receive'. This allows
information to be passed in both directions. For )$M?++ and ")$M?++ the paired
fre4uencies are separated by >A M7(% for @C$<-++ the separation is ?A M7( and for
!C$<?++ separation is -+ M7(.
For each cell in a )$M network at least one 2#FC must be allocated% and more may
be allocated to provide greater capacity. The #F carrier in )$M can support up to eight
Time @ivision Multiple 2ccess &T@M2' timeslots. That is% in theory% each #F carrier is
capable of supporting up to eight simultaneous telephone calls% but as we will see later in
this course although this is possible% network signalling and messaging may reduce the
overall number from eight timeslots per #F carrier to six or seven timeslots per #F
carrier% therefore reducing the number of mobiles that can be supported.
6nlike a !$T network% where every telephone is linked to the land network by a pair of
fixed wires% each M$ only connects to the network over the radio interface when
re4uired. Therefore% it is possible for a single #F carrier to support many more mobile
stations than its eight T@M2 timeslots would lead us to believe. 6sing statistics% it has
been found that a typical #F carrier can support up to <A% 5+ or even 5A M$s. Bbviously%
not all of these M$ subscribers could make a call at the same time% but it is also unlikely
that all the M$ subscribers would want to make a call at the same time. Therefore%
without knowing it% M$s share the same physical resources% but at different times.
FREQUENCY RANGE
GSM 900
#eceive Fre4uency&6plink' C -?+D?<A Mh(
Transmit Fre4uency&@ownlink'0 ?8AD?9+ Mh(
<5> 2bsolute #adio Fre4uency channels
GSM1800(DCS 1800)
#eceive Fre4uency&6plink' C<*<+D<*-A Mh(
Trnamit Fre4uency&@ownlink'0<-+AD<--+ Mh(
8*> 2bsolute #adio Fre4uency channels
PCS 1900
#eceive Fre4uency&6plink' C<-A+D<?<+ Mh(
Transmit Fre4uency&@ownlink'00<?8+D<??+ Mh(
5?? 2bsolute #adio Fre4uency channels
ARFCN
Eandwidth
- timeslots
880 890 915 935 960 925
E-GSM900
UPLINK
GSM900
UPLINK
GSM900
DOWNLIN
K
E-GSM900
DOWNLINK
45 MHz
45 MHz
Guard Band

Channels Assignment
2s shown the 6plink and @ownlink band are separated by 5+ M7( of guard band in case
of )$M and @C$ and <+ M7( in case of "0)$M. The channel separation between
6plink and @ownlink is >A M7( in case of )$M and "0)$M and is ?AM7( in case of
@C$ network. "ach channel&carrier' in )$M system is of 5++ F7( bandwidth% which are
designated by 2bsolute #adio Fre4uency Channel umber &2#FC'. ,f we call Fl&n' the
fre4uency value of the carrier 2#FC n in the lower band&6plink'% and Fu&n' the
corresponding fre4uency value in the upper band &@ownlink'% we have1
GSM 900 Fl(n) 890 + 0.2n 1 <n <124
Fu(n)
Fu(n) Fl(n) +
45
E-GSM 900 Fl(n) 890 + 0.2n
Fl(n) 890 + 0.2(n-1024)
0 <n <124
Fu(n)
975 <n <1023
Fu(n) Fl(n) +
45
DCS 1800 Fl(n) 1710.2 + 0.2(n-512) 512 <n <885 Fu(n) Fl(n) +
95
ARFCN TABLE
7ence we have <5> channels in )$M?++% <*> channels in "0)$M?++ and 8*> channels
in @C$<-++.
FREQUENCY RE-USE
1710
1785
DCS1800
UPLINK
1805
1880
95 MHz
DCS1800
DOWNLINK
Bne important characteristic of )$M networks is fre4uency planning wherein given the
limited fre4uency spectrum available% the re0use of fre4uencies in different cells is to be
planned such that high capacity can be achieved keeping the interference under a specific
level.
2 cell in a )$M system may be omni0directional or sectored represented by hexagons. ,n
)$M system a tri0sectored cell is assumed and the fre4uency plan is made accordingly.
To understand the fre4uency re0use planning% consider a )$M system having S channels
&2#FC;s' allocated% wherein each cell &sector' is allocated k channels% assuming that all
three sectors have same number of k channels. ,f the $ channels are divided among N
base stations each having three sectored cell% then the total number of available radio
channels can be expressed as%
S 3kN
This explains base stations each having three sectors and each sector having k channels
.The N base stations% which collectively use the complete set of available fre4uencies% in
which each fre4uency is used exactly once is called a Cluster. ,f the cluster is replicated
M times then the total number of channels% C% can be used as measure of capacity and is
given by%
C M3kN MS
The Cluster si(e is typically e4ual to 8% >% *% or <5. @eciding a cluster si(e posses a
compromise between capacity% spectrum allocated and interference. 2 cluster si(e of * or
<5 gives least interference fre4uency plan but as the cluster si(e is big enough hence re0
use at far away distance hence lesser capacity and would also re4uire bigger fre4uency
spectrum. Consider an example where k e4uals < that is one fre4uency per sector. .ith a
cluster si(e of * would re4uire minimum spectrum of%
S 3 x 1 x 7 21 ARFCN
or
21 x 0.2 MHz 4.2 MHz
of spectrum that is about 16 of total available spectrum in )$M?++.
2dding one more fre4uency per sector would take the re4uirement to >5 2#FC or 88G
of total spectrum. Bn the other hand a cluster si(e of 8 would re4uire &k H <'%
S 3 x 1 x 3 9 ARFCN
or
9 x 0.2 MHz 1.8 MHZ
which is about 7 of total spectrum available.
2ddition of one more fre4uency still results in about only <>G of spectrum re4uired. Eut
here a big compromise is made on interference% as the cells are 4uite closely located
hence re0use would pose a ma:or problem. $tudies have revealed that cluster si(e of >
gives the best balance between capacity I interference% with k e4ual to 5 meaning two
fre4uencies per sector gives%
S 3 x 2 x 4 24
or
24 x 0.2 MHz 4.8 MHz
that is about 19 of total spectrum available.

Figure illustrates the fre4uency reuse for cluster si(e of >% where cells labelled with the
same letter use the same group of channels.
B1
B2
B3
C1
C2
C3
B1
B2
B3
D1
D2
D3
B1
B2
B3
B1
B2
B3
B1
B2
B3
B1
B2
B3
C1
C2
C3
A1
A2
A3
A1
A2
A3
A1
A2
A3
A1
A2
A3
A1
A2
A3
C1
C2
C3
C1
C2
C3
C1
C2
C3
C1
C2
C3
D1
D2
D3
D1
D2
D3
D1
D2
D3
D1
D2
D3
D1
D2
D3
4 x 3 Re-use pattern
CO-CHANNEL INTERFERENCE AND SYSTEM CAPACITY
Fre4uency re0use implies that in a given coverage area there are several cells that uses the
same set of fre4uencies. These cells are called co0channel cells and the interference
between signals from these cells is called co0channel interference. 6nlike thermal noise
which% can be overcome by increasing the $= ratio% co0channel interference cannot be
combated by simple increase in carrier power. This is because an increase in carrier
power increases the interference to neighbouring co0channel cells. To reduce co0channel
interference% co0channel cells must be physically separated by a minimum distance in
order to provide sufficient isolation due to propagation.

,n a cellular system where the si(e of each cell is approximately the same% co0channel
interference is independent of the transmitted power and becomes the function of the
radius of the cell &#'% and the distance to the centre of the nearest co0channel cell &@'.
Figure explains the relation between the cell radius #% cluster si(e and the re0use
distance @%
7ere%
Outer Cell radius1 R
Inner Cell radius1 r 0.5 x (3)
1/2
x R.
Re-use distance1 D R x (3 x (i
2
+ j
2
+ I j))
1/2
D/R (3 x N )
1/2
The Cluster size% N ( i
2
+ j
2
+ I j )
D
i
j
i
j
D
Re-use distance calculation
.here i and j are non0negative numbers. To find the nearest co0channel neighbour of a
particular cell% one must do the following1
move i cells along any chain of hexagons and then
turn 9+ degrees counter0clockwise and move j cells.

This is illustrated in the figure above for i 1 I j 2 for a cluster si(e of *.

Ey increasing the ratio of @=#% the spatial separation between co0channel cells relative to
the coverage distance of a cell is increased. Thus interference is reduced due to improved
isolation from the co0channel cells. The relation between the re0use distance ratio @=#
and the co0channel interference ratio C=, is as below%

(D/R)

6 (C/I)
&ote1 C=, is in dE and should be converted to numeric values for calculation'
7ere% is the propagation index or attenuation constant with values ranging between 5 to
>.
DESIGN CRITERION
2n optimal fre4uency plan re4uires minimal interference between co0channel and
ad:acent channel cells% )$M #ec. +A.+A has defined the interference ratios for co0channel
and ad:acent channel cells. The actual interference ratio shall be less than a specified
limit% called the reference interference ratio. The reference interference ratio shall be for
base station and all types of M$%

for co channel interference 1 C=, c H ? dE
for ad:acent &5++ k7(' interference 1 C=,a< H 0 ? dE
for ad:acent &>++ k7(' interference 1 C=,a5 H 0 >< dE
For the network planning purpose it is recommended that a value of C=,c ? dE and the
first ad:acent channel C=, a 0? dE. This implies that the first ad:acent channel should not
be used in the same sector cell or the same base station.
r
R

EXAMPLE
2s an illustration let us consider that we re4uire to design a system with C=, of <5 dE and
we have from field drive test results the value of as 8.A% inserting these values in
e4uation
(D/R)

6 (C/I) we have,
(D/R)
3.5
6 x 10.78 64.75
3.5 Log(D/R) Log(64.75) 1.81
(D/R) Antilog(1.81/3.5)
This gives (D/R) 3.29.
.ith this we can back calculate the re4uired cluster si(e from e4uation D/R (3 N)
1/2
as%
H &8.5?'
5
= 8 H 8.9<
7ence a cluster si(e of > will satisfy our re4uired C=, criteria rather if we back calculate
for Cluster of si(e > then we get C=, of <?dE.
FREQUENCY CHANNEL ALLOCATION
,n )$M systems we divide the total allocated spectrum into two sub0groups one for
Control information with traffic referred to as ECC7 fre4uency and other only for traffic
referred to as TC7 &or non0ECC7' fre4uency. ,n case where the network has Microcells
then the total band allotted is divided for ECC7 and TC7% wherein each band is further
sub0divided for Macrocellular I Microcellular applications. Figure &<.8' explains the
concept%
Frequency band allocation.
The re0use may differ for both the groups% as little or no compromise is made for ECC7
fre4uency interference whereas certain compromise could be made for TC7 fre4uency
interference. Typically a cluster si(e of > or * is considered for ECC7 re0use whereas a
TCH TCH
BCCH
Micr C!"" Macr C!""
cluster si(e of 8 or > is used for TC7 re0use. The number of channels in each group
depends on the spectrum allocated and C=, criteria for re0use in each case.
EXAMPLE
2s an example consider C=, criteria of <5 dE for ECC7 then the cluster si(e of > gives
the better result whereas if the C=, criteria is ? dE for TC7% gives the cluster si(e of 8.
The figure &<.<' illustrates the case > x 8 re0use pattern for ECC7 and the figure &<.>'
below illustrates the case of 8 x 8 re0use pattern for TC7%
3 x 3 re-use pattern.
A1 B1 C1 D1 A2 B2 C2 D2 A3 B3 C3 D3
BCC
H
< 5 8 > A 9 * - ? <+ << <5
2< E< C< 25 E5 C5 28 E8 C8
TCH <8 <> <A <9 <* <- <? 5+ 5<
For @C$<-++ planning with cluster si(e of * the fre4uency grouping is as follows%
,n case of @C$<-++ where a large band of spectrum is available the ECC7 and TC7 re0
use can be kept the same.
Set A
1
B
1
C
1
D
1
E
1
F
1
G
1
A
2
B
2
C
2
D
2
E
2
F
2
G
2
A
3
B
3
C
3
D
3
E
3
F
3
G
3
BC
CH
< 5 8 > A 9 * - ? <
+
<
<
<
5
<
8
<
>
<
A
<
9
<
*
<
-
<
?
5
+
5
<
TC
H1
5
5
5
8
5
>
5
A
5
9
5
*
5
-
5
?
8
+
8
<
8
5
8
8
8
>
8
A
8
9
8
*
8
-
8
?
>
+
>
<
>
5
TC
H2
>
8
>
>
>
A
>
9
>
*
>
-
>
?
A
+
A
<
A
5
A
8
A
>
A
A
A
9
A
*
A
-
A
?
9
+
9
<
9
5
9
8
A1
A2
A3
C1
C2
C3
B1
B2
B3
A1
A2
A3
A1
A2
A3
A1
A2
A3
A1
A2
A3
C1
C2
C3
A1
A2
A3
B1
B2
B3
B1
B2
B3
C1
C2
C3
C1
C2
C3
C1
C2
C3
C1
C2
C3
BSIC PLANNING
,n addition to the assignment of fre4uency group to a cell% a Ease $tation ,dentity Code
&E$,C' must be assigned in association with the fre4uency group. This will eliminate the
possibility of incorrect cell identification and will allow the evolution to future cell
architecture. The E$,C is a two0digit code wherein the first digit is indicates CC
&etwork Colour Code' and the second digit indicates ECC &Ease $tation Colour Code'.
The CC and ECC have values ranging from + to *% where the CC is fixed for an
operator% signifying at any given point there can be maximum of - operators in an area.
The ECC defines the cluster number which means a group of - clusters carry uni4ue
identity which are re0used for another group of - clusters and so on. The principal for
allocation of the E$,C is the same as for the #F carriers but at cluster level rather than
cell level. The concept can be understood in the following example%
EXAMPLE
2ssume a network with <++ base stations each having three sectors. The ECC7 and TC7
share the same re0use plan > x 8. .hich means we have cluster of > base stations% and in
all we have <++=> H 5A clusters. 2ssume CC code allocated is 9% which gives us clusters
starting from number 9< to 9*. 7ence seven clusters form a group and hence we have
5A=* that is 8 groups of * clusters plus additional > clusters which form part of the >
th
group. The reuse of these * clusters group for E$,C numbered from 9< to 9* is shown in
the figure &<.A' below1
61
62
63
64
67
61
62
63
64
65
66
61
62
63
64
65
66
67
61
62
63
64
65
66
67
#!$r!%!n& a
c"u%&!r ' ( %i&!%
!ac) )a*in+ ,
%!c&r%
BSIC 7 re-use cluster plan.
,t should be noted that since E$,C are defined at cell &sector' level% hence there are every
possible chances that the three sectors within the same site can have different E$,C. The
reason being as E$,C is used for cell identification hence cells with same ECC7
fre4uency but different E$,C can be easily discriminated by the M$.
AUTOMATIC FREQUENCY PLANNING
2utomatic fre4uency planning is an feature offered by the planning tools to speed up the
work of channel assignment and presents more reliable fre4uency assignment to sites.
2F! &2utomatic fre4uency planning' works on complex algorithm whose calculations
are based on the interference table data% field strength grids and an optional demand
density grid &or traffic distribution table'. ,t allows human interaction at certain points
such as assigning penalties to different clutter types or allowing interference results to be
neglected especially in coverage boundaries of the network. 2F! is of immense help and
provides guidelines in the cases where fre4uency assignment is re4uired for big complex
network. Easic fre4uency planning tool is a standard feature of all available planning
tools% however the advanced 2F! tool based on complex algorithm is provided as an
optional feature.
FREQUENCY HOPPING
The principle of Fre4uency 7opping used within )$M is that successive T@M2 bursts of
a connection are transmitted via different fre4uencies0the fre4uencies belonging to the
respective cell according to network planning. This method is called $low Fre4uency
7opping &$F7' since the transmission fre4uency remains constant during one burst. ,n
contrast to Fast Fre4uency 7opping &FF7' where the transmission fre4uency changes
within one burst.
The effect of fre4uency hopping is that link 4uality may change from burst to burst% ie a
burst of high E"# may be followed by a burst of low E"#% since
$hort term fading is different on different fre4uencies%
The interference level is different on different fre4uencies.
The results of fre4uency hopping are improvement in the received 4uality in fading
situation and interference averaging.
Frequency Hopping Techniques
The hopping techni4ues can be broadly classified into two main categories. They are%
Base band Hopping
Synthesised Hopping
NETWORK ARCHITECTURE
2 cellular telephone system links mobile station &M$' subscribers into the public
telephone system or to another cellular system;s M$ subscriber. ,nformation sent between
the M$ subscriber and the cellular network uses radio communication. This removes the
necessity for the fixed wiring used in a traditional telephone installation.
@ue to this% the M$ subscriber is able to move around and become fully mobile% perhaps
travelling in a vehicle or on foot.

NETWORK COMPONENTS
)$M networks are made up of Mobile services $witching Centres &M$C'% Ease $tation
$ystems &E$$'and Mobile $tations &M$'. These three entities can be broken down
further into smaller entities/ such as% within the E$$ we have Ease $tation Controllers%
Ease Transceiver $tations and Transcoders.
These smaller network elements% as they are referred to% will be discussed later in the
course. For now we will use the three ma:or entities. .ith the M$C% E$$ and M$ we can
make calls% receive calls% perform billing etc% as any normal !$T network would be able
to do. The only problem for the M$ is that not all the calls made or received are from
other M$s. Therefore% it is also necessary to connect the )$M network to the !$T.
Mobile $tations within the cellular network are located in cells% these cells are provided
by the E$$s. "ach E$$ can provide one or more cells% dependent on the manufacturers
e4uipment.
The cells are normally represented by a hexagon% but in practice they are irregular in
shape. This is as a result of the influence of the surrounding terrain% or of design by the
network planners.
.
SWITCHING AND CONTROL
7aving established radio coverage through the use of cells% both omni0directional and
directional &sectored sites'% now consider what happens when the M$ is in motion &as
M$s tend to be'.
2t some point the M$ will have to move from one cell;s coverage area to another cell;s
coverage area. 7andovers from one cell to another could be for a number of reasons
&e.g. the signal strength of the serving cell is less than the signal strength of a
neighbour cell% or the M$ is suffering a 4uality problem in the serving cell' and by
handing over to one of its neighbours this may stop the 4uality problem.
#egardless of the reason for a handover it has to be controlled by some entity% and in
)$M that entity is the Mobile services $witching Centre &M$C'.
To perform a handover% the network must know which neighbour cell to hand the M$
over to. To ensure that we handover to the best possible candidate the M$ performs
measurements of its surrounding neighbour cells and reports its findings to the network.
These are then analy(ed together with the measurements that the network performs and
a decision is made on a regular basis as to the need for a handover. ,f a handover is
re4uired then the relevant signal protocols are established and the handover is controlled
by the M$C.
7andovers must be transparent to the M$ subscriber. That is the subscriber should be
unaware that a handover has occurred.
2s we will see later in this course% handovers are :ust one of the functions of the M$C%
many more are performed by the M$C and its associated entities &e.g. such as
authentication of M$% ciphering control% location updating% gateway to !$T'.
Note
$ome networks may allow certain handovers to be performed at the E$$ level. This
would be dependent on the manufacturer;s e4uipment.
Switching and Control
GSM NETWORK OVERVIEW
The diagram shows a simplified )$M network. "ach network component is
illustrated only once% however% many of the components will occur several times
throughout a network.
"ach network component is designed to communicate over an interface specified by the
)$M standards. This provides flexibility and enables a network provider to utili(e system
components from different manufacturers. For example Motorola Ease $tation $ystem
&E$$' e4uipment may be coupled with an "ricsson etwork $witching $ystem.
The principle component groups of a )$M network are1
The Mobile Station (MS)
This consists of the mobile telephone% fax machine etc. This is the part of the
network that the subscriber will see.
The Base Station System (BSS)
This is the part of the network which provides the radio interconnection from the
M$ to the land0based switching e4uipment.
The Network Switching System
This consists of the Mobile services $witching Centre &M$C' and its associated
system0control databases and processors together with the re4uired interfaces.
This is the part which provides for interconnection between the )$M network and
the !ublic $witched Telephone etwork &!$T'.
The Operations and Maintenance System
This enables the network provider to configure and maintain the network from a
central location.
GSM NETWORK COMPONENTS
MOBILE STATION (MS)
The M$ consists of two parts% the Mobile "4uipment &M"' and an electronic Jsmart card`
called a $ubscriber ,dentity module &$,M'.
The M" is the hardware used by the subscriber to access the network. The hardware
has an identity number associated with it% which is uni4ue for that particular device and
permanently stored in it. This identity number is called the ,nternational Mobile
"4uipment ,dentity &,M",' and enables the network operator to identify mobile
e4uipment
which may be causing problems on the system.
The $,M is a card which plugs into the M". This card identifies the M$ subscriber and
also provides other information regarding the service that subscriber should receive. The
subscriber is identified by an identity number called the ,nternational Mobile $ubscriber
,dentity &,M$,'.
Mobile "4uipment may be purchased from any store but the $,M must be obtained from
the )$M network provider. .ithout the $,M inserted% the M" will only be able to make
emergency calls.
Ey making a distinction between the subscriber identity and the M" identity% )$M can
route calls and perform billing based on the identity of the Jsubscriber; rather than the
e4uipment or its location.

MOBILE EQUIPMENT (ME)

The M" is the only part of the )$M network which the subscriber will really see. There
are three main types of M"% these are listed below1
Vehicle Mounted
These devices are mounted in a vehicle and the antenna is physically mounted on
the outside of the vehicle
.
Portable Mobile Unit
This e4uipment can be handheld when in operation% but the antenna is not
connected to the handset of the unit.
Hand portable Unit
This e4uipment comprises of a small telephone handset not much bigger than a
calculator. The antenna is be connected to the handset.
The M" is capable of operating at a certain maximum power output dependent on its
type and use.
These mobile types have distinct features which must be known by the network% for
example their maximum transmission power and the services they support. The M" is
therefore identified by means of a class mark. The class mark is sent by the M" in its
initial message.
The following pieces of information are held in the classmark1

Mobile equipment capabilies:

#F power Capability
$upport of phase< %!hase5 or !hase5K specifications
"ncryption capability
Fre4uency capability
$hort message services capability

SUBSCRIBER IDENTITY MODULE (SIM)

The $,M as mentioned previously is a smart card which plugs into the M" and contains
information about the M$ subscriber hence the name $ubscriber ,dentity Module.
The $,M contains several pieces of information1
International Mobile Subscriber Identity (IMSI)
This number identifies the M$ subscriber. ,t is only transmitted over the air during
initiali(ation.
Temporary Mobile Subscriber Identity (TMSI)
This number identifies the subscriber% it is periodically changed by the system
management to protect the subscriber from being identified by someone
attempting to monitor the radio interface.
Location Area Identity (LAI)
,dentifies the current location of the subscriber.
Subscriber Authentication Key (Ki)
This is used to authenticate the $,M card.
Mobile Station International Services Digital Network (MSISDN)
This is the telephone number of the mobile subscriber. ,t is comprised of a country
code% a network code and a subscriber number.
Most of the data contained within the $,M is protected against reading &Fi' or alterations
&,M$,'. $ome of the parameters &32,' will be continuously updated to reflect the current
location of the subscriber.
The $,M card% and the high degree of inbuilt system security% provide protection of the
subscriber;s information and protection of networks against fraudulent access. $,M
cards are designed to be difficult to duplicate. The $,M can be protected by use of
!ersonal ,dentity umber &!,' password% similar to bank=credit charge cards% to prevent
unauthori(ed use of the card.
The $,M is capable of storing additional information such as accumulated call charges.
This information will be accessible to the customer via handset=keyboard key entry.
The $,M also executes the 2uthentication 2lgorithm.
BASE STATION SYSTEM (BSS)
The )$M Ease $tation $ystem is the e4uipment located at a cell site. ,t comprises a
combination of digital and #F e4uipment. The E$$ provides the link between the M$
and the M$C.
The E$$ communicates with the M$ over the digital air interface and with the M$C via
5 Mbit=s links.
The E$$ consists of three ma:or hardware components1
The Base Transceiver Station - BTS
The ET$ contains the #F components that provide the air interface for a particular
cell. This is the part of the )$M network which communicates with the M$. The
antenna is included as part of the ET$.
The Base Station Controller - BSC
The E$C as its name implies provides the control for the E$$. The E$C
communicates directly with the M$C. The E$C may control single or multiple
ET$s.
The Transcoder - XCDR
The Transcoder is used to compact the signals from the M$ so that they are
more efficiently sent over the terrestrial interfaces. 2lthough the transcoder is
considered to be a part of the E$$% it is very often located closer to the M$C.
The transcoder is used to reduce the rate at which the traffic &voice=data' is transmitted
over the air interface. 2lthough the transcoder is part of the E$$% it is often found
physically closer to the $$ to allow more efficient use of the terrestrial links.
Base Station Controller (BSC)
2s previously mentioned% the E$C provides the control for the E$$. The functions of
the E$C are shown in the table opposite.
2ny operational information re4uired by the ET$ will be received via the E$C. 3ikewise
any information re4uired about the ET$ &by the BMC for example' will be obtained by
the
E$C.
The E$C incorporates a digital switching matrix% which it uses to connect the radio
channels on the air interface with the terrestrial circuits from the M$C.
The E$C witching matrix also allows the E$C to perform handovers between radio
channels on ET$s% under its control% without involving the M$C.
Base Transceiver
Station - BTS
The ET$ provides the air interface connection with the M$. , also has a limited amount
of
control functionality which reduces the amount of traffic passing between the ET$ and
E$C. The functions of the ET$ are shown opposite. "ach ET$ will support < or more
cells.
.here the E$C and ET$ are both shown to control a function% the control is divided
between the two% or may be located wholly at one.
BSS CONFIGURATIONS
2s we have mentioned% a E$C may control several ET$s% the maximum number of ET$s
which may be controlled by one E$C is not specified by )$M.
,ndividual manufacturer;s specifications may vary greatly.
The ET$s and E$C may either be located at the same cell site co0located, or located
at different sites #emote. ,n reality most ET$s will be remote% as there are many more
ET$s than E$Cs in a network.
2nother E$$ configuration is the daisy chain. 2 ET$ need not communicate directly with
the E$C which controls it% it can be connected to the E$C via a chain of ET$s.
@aisy chaining reduces the amount of cabling re4uired to set up a network as a ET$ can
be connected to its nearest ET$ rather than all the way to the E$C.
!roblems may arise when chaining ET$s% due to the transmission delay through the
chain. The length of the chain must% therefore% be kept sufficiently short to prevent the
round trip speech delay becoming too long.
Bther topologies are also permitted% including stars and loops. 3oops are used to
introduce redundancy into the network% for example if a ET$ connection was lost% the
ET$ may still be able to communicate with the E$C if a second connection is available.
TRANSCODER (XCDR)
The Transcoder &LC@#' is re4uired to convert the speech or data output from the M$C
&9> kbit=s !CM'% into the form specified by )$M specifications for transmission over the
air interface% that is% between the E$$ and M$ &9> kbit=s to <9 kbit=s and vice versa'.
The 9> kbit=s !ulse Code Modulation &!CM' circuits from the M$C% if transmitted on the
air interface without modification% would occupy an excessive amount of radio
bandwidth.
This would use the available radio spectrum inefficiently. The re4uired bandwidth is
therefore reduced by processing the 9> kbit=s circuits so that the amount of information
re4uired to transmit digiti(ed voice falls to a gross rate of <9 kbit=s.
The transcoding function may be located at the M$C% E$C% or ET$.
The content of the <9 kbit=s data depends on the coding algorithm used. There are two
speech coding algorithms available and selecting which one to use depends on the
capabilities of the mobile e4uipment and the network configuration.
The Full #ate speech algorithm is supported by all mobiles and networks. ,t produces <8
kbit=s of coded speech data plus 8 kbit=s of control data which is commonly referred to as
T#26 data &Transcoder #ate 2daption 6nit'. The T#26 data on the downlink will be
used by the ET$ and therefore removed from the <8 k of speech data before
transmission on the air interface. the <8 kbit=s of speech data is processed at the ET$ to
form a gross rate of 55.- kbit=s on the air interface which includes forward error
correction. ,n the uplink direction the ET$ adds in T#26 data which will be used by the
transcoder.
"nhanced Full #ate is an improved speech coding algorithm and is only supported by
!hase 5K mobiles and is optional in the etwork. ,t produces <5.5 kbit=s from each 9>
kbit=s !CM channel. The T#26 data in this case is made up to 8.- kbit=s to keep the
channel rate to and from the ET$ at <9 kbit=s as for Full #ate. 2s with Full #ate the
T#26 data is used at the ET$ and Transcoder.
For data transmissions the data is not transcoded but data rate adapted from ?.9 kbit=s
&>.- kbit=s or 5.> kbit=s may also be used' up to a gross rate of <9 kbit=s for transmission
over the terrestrial interfaces% again this <9 kbit=s contains a 8 kbit=s T#26.
2s can be seen from the diagram opposite% although the reason for transcoding was to
reduce the data rate over the air interface% the number of terrestrial links is also reduced
approximately on a >1< ratio.
NETWORK SWITCHING SYSTEM
The etwork $witching $ystem includes the main switching functions of the )$M
network. ,t also contains the databases re4uired for subscriber data and mobility
management. ,ts main function is to manage communications between the )$M
network and other telecommunications networks.
The components of the etwork $witching $ystem are listed below1
Mobile $ervices $witching Centre C MSC
7ome 3ocation #egister C HLR
Misitor 3ocation #egister C VLR
"4uipment ,dentity #egister C EIR
2uthentication Centre C AUC
,nter.orking Function C IWF
"cho Canceller C EC
,n addition to the more traditional elements of a cellular telephone system% )$M has
3ocation #egister network entities. These entities are the 7ome 3ocation #egister
&73#'% Misitor 3ocation #egister &M3#)% and the "4uipment ,dentity #egister &",#'. The
location registers are database0oriented processing nodes which address the problems of
managing subscriber data and keeping track of a M$s location as it roams around the
network.
Functionally% the ,nterworking Function and the "cho Cancellers may be considered as
parts of the M$C% since their activities are inextricably linked with those of the switch as
it
connects speech and data calls to and from the M$s.

MOBILE SERVICES SWITCHING CENTRE (MSC)
The M$C is included in the )$M system for call0switching. ,ts overall purpose is the
same as that of any telephone exchange.
7owever% because of the additional complications involved in the control and security
aspects of the )$M cellular system and the wide range of subscriber facilities that it
offers% the M$C has to be capable of fulfilling many additional functions.
The M$C will carry out several different functions depending upon its position in the
network. .hen the M$C provides the interface between the !$T and the E$$s in the
)$M network it will be known as a )ateway M$C. ,n this position it will provide the
switching re4uired for all M$ originated or terminated traffic.
"ach M$C provides service to M$s located within a defined geographic coverage area%
the network typically contains more than one M$C. Bne M$C is capable of supporting a
regional capital with approximately one million inhabitants. 2n M$C of this si(e will be
contained in about half a do(en racks.
The functions carried out by the M$C are listed below1
Call Processing
,ncludes control of data=voice call setup% inter0E$$ and inter0M$C handovers and
control of mobility management &subscriber validation and location'.
Operations and Maintenance Support
,ncludes database management% traffic metering and measurement% and a
manCmachine interface.
Internetwork Interworking
Manages the interface between the )$M network and the !$T.
Billing
Collects call billing data.
HOME LOCATION REGISTER (HLR)
The 73# is the reference database for subscriber parameters.
Marious identification numbers and addresses are stored% as well as authentication
parameters. This information is entered into the database by the network provider when
a new subscriber is added to the system.
The parameters stored in the 73# are listed opposite1
The 73# database contains the master database of all the subscribers to a )$M !3M.
The data it contains is remotely accessed by all the M$Cs and the M3#s in the network
and % although the network may contain more than one 73#% there is only one database
record per subscriber 0 each 73# is therefore handling a portion of the total subscriber
database. The subscriber data may be accessed by either the ,M$, or the M$,$@
number. The data can also be accessed by an M$C or a M3# in a different !3M% to
allow inter0system and inter0country roaming.
VISITOR LOCATION REGISTER (VLR)
The M3# contains a copy of most of the data stored at the 73#. ,t is% however%
temporary data which exists for only as long as the subscriber is active in the particular
area covered by the M3#. The M3# database will therefore contain some duplicate data
as well as more precise data relevant to the subscriber remaining within the M3#
coverage.
The M3# provides a local database for the subscribers wherever they are physically
located within a !3M% this may or may not be the home system. This function
eliminates the need for excessive and time0consuming references to the home 73#
database.
The additional data stored in the M3# is listed below1
Mobile status &busy=free=no answer etc.'.
Location Area Identity &32,'.
Temporary Mobile Subscriber Identity &TM$,'.
Mobile Station Roaming Number &M$#'
EQUIPMENT IDENTITY REGISTER (EIR)
The ",# contains a centrali(ed database for validating the ,nternational Mobile
"4uipment ,dentity &,M",'.
This database is concerned solely with M$ e4uipment and not with the subscriber who is
using it to make or receive a call.
The ",# database consists of lists of ,M",s &or ranges of ,M",s' organi(ed as follows1
White List
Contains those ,M",s which are known to have been assigned to valid M$
e4uipment.
Black List
Contains ,M",s of M$ which have been reported stolen or which are to be denied
service for some other reason.
Grey List
Contains ,M",s of M$ which have problems &for example% faulty software'. These
are not% however% sufficiently significant to warrant a JJblack listing.
The ",# database is remotely accessed by the M$Cs in the network and can also be
accessed by an M$C in a different !3M.
2s in the case of the 73#% a network may well contain more than one ",# with each ",#
controlling certain blocks of ,M", numbers. The M$C contains a translation facility%
which
when given an ,M",% returns the address of the ",# controlling the appropriate section of
the e4uipment database.
OPERATIONS AND MAINTENANCE SYSTEM
The operations and maintenance system provides the capability to manage the )$M
network remotely.
This area of the )$M network is not currently tightly specified by the )$M
specifications%
it is left to the network provider to decide what capabilities they wish it to have. The
Bperations and Maintenance $ystem comprises of two parts1
NETWORK
MANAGEMENT
CENTRE (NMC)
The etwork Management Centre &MC) has a view of the entire !3M and is
responsible for the management of the network as a whole. The MC resides at the top
of the hierarchy and provides global network management.
OPERATIONS AND
MAINTENANCE
CENTRE (OMC)
The Bperations and Maintenance Centre &BMC' is a centrali(ed facility that supports the
day to day management of a cellular network as well as providing a database for long
term network engineering and planning tools. 2n BMC manages a certain area of the
!3M thus giving regionali(ed network management.
NETWORK MANAGEMENT CENTRE (NMC)
The MC offers the ability to provide hierarchical regionali(ed network management of
a
complete )$M system.
,t is responsible for operations and maintenance at the network level% supported by the
BMCs which are responsible for regional network management.
The MC is therefore a single logical facility at the top of the network management
hierarchy.
The MC has a high level view of the network% as a series of network nodes and
interconnecting communications facilities. The BMC% on the other hand% is used to filter
information from the network e4uipment for forwarding to the MC% thus allowing it to
focus on issues re4uiring national co0ordination. The MC can also co0ordinate issues
regarding interconnection to other networks% for example the !$T.
The MC can take regional responsibility when an BMC is not manned% with the BMC
acting as a transit point between the MC and the network e4uipment. The MC
provides operators with functions e4uivalent to those available at the BMC.
OPERATIONS AND MAINTENANCE CENTRE (OMC)
The BMC provides a central point from which to control and monitor the other network
entities &i.e. base stations% switches% database% etc' as well as monitor the 4uality of
service being provided by the network.
2t present% e4uipment manufacturers have their own BMCs which are not compatible in
every aspect with those of other manufacturers. This is particularly the case between
radio base station e4uipment suppliers% where in some cases the BMC is a separate
item and @igital $witching e4uipment suppliers% where the BMC is an integral% but
functionally separate% part of the hardware.
There are two types of BMC these are1
OMC (R)
BMC controls specifically the Ease $tation $ystem.
OMC (S)
BMC controls specifically the etwork $witching $ystem.
The BMC should support the following functions as per ,T$CT$ recommendations1
"vent=2larm Management.
Fault Management.
!erformance Management.
Configuration Management.
$ecurity Management.
The BMC functional architecture is illustrated in the diagram opposite.
OMC Functional Architecture
THE NETWORK IN REALITY
,n reality a )$M network is much more complicated than we have seen. The diagram
opposite illustrates how multiple E$$ and etwork $witching $ystem components will
be
connected within a network.
2 typical city for example% 3ondon' will have approximately the following number of
network components1