Technical and economical assessment of selected LTE-A

schemes.
Heinz Droste, Deutsche Telekom Laboratories, Darmstadt
Project Field „Intelligent Wireless Technologies & Networks“

1
Mobile Networks – enabler for connected life & work.

Textbox Headline
Customer demands Technology
Textbox Headline
evolution
 Mobility, flexibility and convenience are key  Fixed and wireless access technologies are
user requirements for access to Internet-based constantly evolving.
services on the move, at work and at home  Date rate capabilities double about every
 can only be provided by wireless access. 18 months.
 Data rate requirements for mobile and wireless  Data rates of mobile technologies are about a
access are rapidly increasing in future. factor 10 … 25 below fixed technologies*.
MBit/s
1,000

Average data rate per user

100
olution
ed ev
10
Fi x tion
l u
i l e evo
Mob
1

0.1

0.01
2006 2009 2012 2015
*) WLAN (Wireless Local Area Networks) support similar data
rates than fixed access).

Deutsche Telekom Laboratories 2

Exponential growth of data traffic in mobile networks.

Textbox Headline
Demand for future mobile communications markets and services in Europe - example.

Typical 10 000 base

station network
deployment assumed.

Source: Analysys Research, 2008.

Deutsche Telekom Laboratories 3

Expected roadmap of mobile network technologies.

2000 - 2003 2003 - 2004 2005 - 2006 2007 - 2010 2011+ 2015+
32
32 -- 128
128 kbps
kbps 64
64 -- 384
384 kbps
kbps 0.384
0.384 -- 44 Mbps
Mbps 0.384
0.384 -- 77 Mbps
Mbps up
up to
to 170
170 Mbps
Mbps (peak)
(peak) up
up to
to 11 Gbps
Gbps

GSM
GSM UMTS
UMTS HSDPA
HSDPA HSPA
HSPA // HSPA+
HSPA+ NGMN -- LTE
NGMN LTE LTE-Advanced
LTE-Advanced
GPRS/
GPRS/ R99,
R99, 1st
1st version
version Downlink
Downlink Downlink
Downlink // Uplink
Uplink Broadband radio,
Broadband radio, Enhanced
Enhanced LTE
LTE with
with
EDGE
EDGE of 3G
of 3G Enhanced
Enhanced 3G
3G Enhanced
Enhanced 3G3G IP based architecture
IP based architecture new radio concepts
new radio concepts
Future Wireless
Future Wireless Cellular
Cellular

Enhanced Multimedia Enhanced Optimized Broadband Mobile

Mobile Services Cellular Multimedia Mobile Multimedia Mobile Communication

NGMN - LTE LTE-Advanced

Optimized UMTS
Enhanced UMTS Scope of T-Labs
3G - UMTS
Innovation Projects

GSM (GPRS / EDGE)

Deutsche Telekom Laboratories 4

Future challenges for mobile network operators.

Higher capacity includes increase of spectrum efficiency.

Source: Dr. K.-J. Krath, Workshop LTE Evolution,

June 15th 2009, Dresden.

Deutsche Telekom Laboratories 5

Spectral efficiency can be increased – but what are the costs?

Arrangetrade-off
Arrange trade-off between
between technical
technical performance
performance & economic
& economic feasibility
feasibility

Example: Uplink network MIMO  Vision & theory: If channel conditions of all
mobiles were available in the whole network
12
simultaneously, transmission capability could
UPLINK 10 be increased enormously.
10 spectral efficiency
bit/s/Hz/sector
8
 Practical constraints: If at all (from a current
perspective), this can only be achieved using
6
Trade- very expensive high-performance networking
4
off? between base station sites.

2 1,5
0,75  EASY-C economic assessment will help find a
0,1 0,18
reasonable trade-off between affordable
0
network components / implementation and
GSM HSPA UL LTE EASY-C Theory
feasible performance gain.
Target

Deutsche Telekom Laboratories 6

Key technologies for LTE-Advanced.

Coordinated Multi-Point (CoMP): Principles and implications

Textbox Headline

 Inter-cell interference mitigation due to cooperation of

adjacent base station ( distributed MIMO system).
 Basic CoMP types:
 Joint processing (“Network MIMO”):
Coherent joint transmission/reception from/to
geographically separated antennas.
 Coordinated scheduling and/or beamforming:
Decisions for scheduling/selection of beams considers
interference situation in neighboring cells.
 CoMP has high potential for big improvement of
spectral efficiency.
 (Fully) meshed broadband backbone necessary with
tight latency and synchronization requirements.
 Backbone capacity and/or latency may impact
achievable gains.

Deutsche Telekom Laboratories 7

Key technologies for LTE-Advanced.
Technologies selected by DTAG for Performance Evaluation

LTE-A Technologies

Uplink Downlink

Base Station Advanced Transmission

Cooperation Schemes

Cooperative Cooperative
Multiuser MIMO
Scheduling Signal Processing

Interference Joint Detection & Multiuser MIMO

Joint Scheduling Distributed SIC
Prediction Link Adaptation 4x2 (CSI)

SIC = Successive Interference Cancellation; CSI = Channel State Information

Deutsche Telekom Laboratories 8

Performance improvements and backhaul requirements.
Results of System Level Simulations

Spectral Cell-to-cell (X2) Throughput

Efficiency Gain[%] Requirement [Mbps]
70 350 340
Investigated Schemes
60 4 300
1UL Interference Prediction
50 6 250
Network capacity

2 UL Joint Scheduling
200
40 3UL Inter-site Joint Detection 200
3
30 3 3 UL Intra-site Joint Detection (red) 150
1
20 4 UL Distributed SIC 100
60
10 5 2 5 DL MU MIMO CSI based 4x2 50
6 UL Intra-site JD & Inter-site IP 8 8
0 0
0 10 20 30 40 50 60 70 80 90 1 2 3 4 6
Cell Edge Throughput Gain [%] LTE-A Scheme
Notes:
User experience Gains relative to LTE Rel. 8
X2 throughput for decentralized approach (no aggregation)

SIC = Successive Interference Cancellation; CSI = Channel State Information; JD = Joint Detection; IP = Interference Prediction

Deutsche Telekom Laboratories 9

Approach for Economic Assessment of LTE-A key technologies.
SON
SON Applicability
Applicability

Baseline
Baseline Baseline
Baseline LTE-A
LTE-A LTE-A
LTE-A
Deployment
Deployment Performance
Performance Deployment
Deployment Performance
Performance

 Economical
Economical Scenario
Scenario  Technical
Technical Scenario
Scenario  Definition
Definition of
of cost
cost factors
factors for
for  System
System level
level simulation
simulation
Definition
Definition Definition
Definition all
all elements
elements except
except for
for results
results for
for spectral
spectral
Backhaul System
System level
level simulations
simulations backhaul
backhaul efficiency
efficiency and
and cell
cell edge
edge
 Backhaul modeling
modeling 
user
Deployment
based
based on 3GPP simulation
on 3GPP simulation  S1&X2
S1&X2 backhaul
backhaul modeling
modeling user throughput
throughput
 Deployment simulation
simulation conditions
conditions


Increased
 Cost per bit Increased
Deployment
Deployment
reduction



Costs
Costs
Improved
Improved
factor



Performance
Performance



    

   
Comparison of rel. cost and
Economic Evaluation Technical performance figures
Scenario Evaluation Scenario
SON = Self Organizing Networks

Deutsche Telekom Laboratories 10

Economic Assessment of LTE-A key technologies.
Cost are estimated by “Deployment Simulations”
Textbox Headline

Real World  Deployment scenarios should allow for detection of all significant topics in
Scenarios context with deployment.
 Mean inter-site distances is chosen as the most significant selection parameter.

Backhaul Models  Backhaul models should allow for estimation of the cost impact of LTE-A
throughput and latency requirements.
 Due to big differences of real world scenarios abstractions are required.

Cost Models  Assessment is oriented towards relative cost figures.

 The relationships between the different cost elements shall be as realistic as
possible.

Clustering Rules  Assumptions on rules for building of CoMP cooperation clusters must be
aligned between performance evaluation and deployment scenarios.
 Cluster building is based on received power levels as well as on signal arriving
times at the receiver (guard interval violation not allowed).

Deutsche Telekom Laboratories 11

Economic Assessment of LTE-A key technologies.
Performance improvements versus cost increase  “Cost per bit Reduction”
Cost per bit Reduction [%]
30
Increased
Increased
Deployment
Deployment
6
Cost per bit Costs 25
Improved Costs
Improved
Spectral
Reduction 4
Spectral 20
Efficiency
Efficiency
3
15
1
10 3
2
5
Cost elements; example: UL Distributed SIC 5
80,00
0
0 10 20 30 40 50 60 70
Difference [%]

60,00

40,00 Spectral Efficiency Gain [%]

20,00

0,00 Ranking according to “Cost per bit Reduction”

6 UL Intra-site JD & Inter-site IP 3 UL Inter-site Joint Detection
E
ul
EX

l
t
EX

ta
en

4 UL Distributed SIC 5 DL MU MIMO CSI based 4x2

M
tio

se

ha

To
AP

M
dm
AP

ra

es

ck
pe
C

3UL Intra-site Joint Detection (red) 2 UL Joint Scheduling

oc

Ba
C

en

O
a

Pr
eB

Am
nn

1UL Interference Prediction

od

te

te
An
eN

Si

SIC = Successive Interference Cancellation; CSI = Channel State Information; JD = Joint Detection; IP = Interference Prediction

Deutsche Telekom Laboratories 12

Thank you for your attention.

Deutsche Telekom Laboratories 13