S

System AAspects off

LTE
Overview

• Motivation for
f 3GPP LTE
• Requirements
• Enabling
bl Technologies
h l
• LTE versus WiMAX
• LTE-Advanced
Ad d
• Selg Organized Networks
• C l
Conclusion
 What is LTE ?

• In Nov.
No 2004
2004, 3GPP began a project to define the long-term
long term
evolution (LTE) of Universal Mobile Telecommunications
S t (UMTS) cellular
System ll l ttechnology
h l
– Higher performance
– Cost Effective
– Wide application
– Does not have to be fully compatible to 2G and 3G
 Wireless World Map

System Application Distance Mobility Data rate

ZigBee Industrial Short Low Low (< 1 Mbps)

Bluetooth PC, Cell phone Short Low Medium ( 1 Mbps)

UWB Home appliances Short Low High ( > 50 Mbps)

WiFi PC network Medium Low High (> 100 Mbps)

WiMAX fixed Internet Service Long Low Medium

2/2.5 G Cellular Voice + low rate data Long Medium Low

3/3 5G Cellular
3/3.5G C ll l V i + Internet
Voice It t L
Long Hi h
High M di
Medium

4G Cellular WiMAX & LTE Voice + Internet +?? Long High High

4G Advanced Voice + Internet + ?? + ?? Long High Very high ( > 200 Mbps)

4
Wireless Map

5
 IEEE and Wireless Systems

• IEEE members produced several wireless standards

6
Expected LTE Subscribers

7
Historical Background

8
Data Rates of Old Systems

Generation Peak bps

p Average
g bps
p Technology
gy

GPRS 115 k 30-40 k TDMA

EDGE 473 k 100-130 k TDMA

UMTS 2M 220-320 k CDMA

UMTS-HSDPA 3.6 & 14 M 550-1100 k CDMA

CDMA-2000 1X 153 k 50-70 k CDMA

CDMA-2000 1XEV-DO 2.4 M 300-500 k CDMA

CDMA-2000 1XEV-DV 3.09 M >1M CDMA

9
Why 4G ?

Current 4G
Voice communication VoIP, high quality video conferencing
SMS MMS
SMS, Vid messaging
Video i
Internet browsing Super-fast internet
Downloadable games Online gaming with mobility
Downloadable video High quality audio & video streaming
No TV service Broadcast TV on-demand
Peer-to-peer
Peer to peer messaging Wide-scale
Wide scale distribution of video clips
Mobile payment
File transfer
Many other innovative ideas

10
 Technical Requirements

• Increase data rate

– About 100 Mbps downlink and 50 Mbps uplink
• Decrease Latency
• Improve wireless performance
– Better signal reception and better coverage
• Increase spectrum efficiency
– More subscribers and more data transfer in the same spectrum
p
• High flexibility of allocation
– Quickly adjust data rate to subscriber according to need
• Spectrum
p flexibilityy
– Several bandwidths can be used
• Optimized for low speed
– Best at < 20 km/hr & support up to 350 km/hr
• Add Multi-Cast and Broadcast Services
– To support broadcast services like TV
• Faster call setup

11
 4G Enabling Technology

• Some key technologies made 4G possible

• Both WiMAX and LTE use:
– OFDM,
OFDM OFDMA and SC SC-FDMA
FDMA
– Channel dependent scheduling
– Adaptive coding and modulation (ACM)
– Multiple-In-Multiple-Out (MIMO) antenna processing
– Turbo coding and decoding
• Need to fight the fading channel

12
 Wireless Fading Channels

• The wireless channel is subject to multipath

(reflections)
• The received signal
g is affectingg byy fadingg

• Fading is a variation in the received signal level

level, with more lows than
highs
• Both the time domain and frequency domain are affected

13
 Fading Channels in Time

• Channel gain changes with time

• In most cases the changes are slower than data rates
• This is not good since consecutive stream of bits may be lost

14
 Fading Channels in Frequency

• Channel gain varies with frequency

• May cause distortion to the signal spectrum, depending on signal bandwidth

• Problem increases with larger

g
bandwidth from 2G to 4G

15
 OFDM Concept

• We have a high rate (hence, large bandwidth) stream of modulation

symbols Xk (ex. QAM)
• Stream
St Xk isi di
divided t N low
id d iinto l rate
t parallel
ll l sub-streams
b t
• Bandwidth of each sub-stream is N times narrower
• Each sub-stream is carried by one subcarrier
• RReceived t eachh Xk without
i d mustt restore ith t iinterference
t f from
f currentt or
previously transmitted sub-streams
• Needs to be transmitted on a frequency selective fading channel

16
LTE-Downlink (OFDM)
( )

• Improved
p oved spec
spectrala eefficiency
c e cy
• Reduce ISI effect by multipath
• Against frequency selective fading
 OFDM Concept
• Transmitted OFDM Signal

• Received OFDM Signal

18
OFDM Concept

e j0

e j1 1 N 1
 2 kn 
xn 
N
X
k 0
k exp 
 N 


e jN 1

19
 OFDM Concept

• OFDM modulation
d l ti using
i IFFT
• Guard time (cyclic prefix) is added to protect against inter-symbol
interference
• Guard subcarriers to pprotect against
g neighbor
g channels at both sides
• Zero subcarrier (dc) not used
• Some b i are usedd as pilots
S subcarriers il for
f channel
h l estimation
i i
• After equalization, receiver performs FFT to retrieve back the stream
Xk

20
OFDM Concept

21
 OFDMA Concept

• In OFDM one user occupies all subcarriers all the time (till packet is finished)
• In OFDMA each user occupies few subcarriers for few OFDM symbols during a Burst of transmission
• A Burst: few subcarriers during few OFDM symbols
• Hence the name Orthogonal Frequency Division Multiplex Access

22
 OFDMA Flexibility

• With OFDMA the user allocation is flexible

– Can change from frame to frame
– Multiple allocations for several applications
• Allocation
All i changes
h
– In WiMAX every 5 ms
– In LTE every 1 ms

23
OFDM Characteristics
h

• Preservation of orthogonality in severe multi-path

• Efficient FFT based receiver structures
• Enables efficient TX and RX diversity
• Adaptive antenna arrays without joint equalization
• Support for adaptive modulation by subcarrier
• Frequency diversity
• Robust against narrow-band interference
• Variable/dynamic bandwidth
• Used for highest speed applications
• Supports dynamic packet access
• High peak-to-average power levels
 Si l CCarrier
Single i FDMA (SC
(SC-FDMA)
FDMA)

• A major problem with OFDM and OFDMA is high peak peak-to-average

to average
power ratio (PAPR)
– Transmitted
T itt d amplitude
lit d with
ith llarge variation
i ti
– Requires a linear amplifier at transmitter
– Linear amplifies consumes high power
– OK at base station
– For mobile station, this consumes battery
• LTE uses a solution for UL: SC-FDMA
– Single carrier transmission

25
 SC FDMA Process
SC-FDMA

• After modulation,
modulation apply FFT
– Each symbol is on a subcarrier
• Put the subcarriers on selected location and apply IFFT
– Back to single carrier transmission
– Now add CP
• Receiver will do the reverse

26
 LTE Uplink (SC-FDMA)

• SC-FDMA is a new single carrier multiple access technique which has similar structure and performance to OFDMA

A salient advantage
of SC-FDMA over
OFDM is low to
Peak to Average
Power Ratio
(PAPR) :
Increasing
battery life
PAPR reduction in SC
SC-FDMA
FDMA

28
 Channel Dependent Scheduling

• Another benefit for OFDMA

• The Base Station talks to many users at different directions
• Their fading channels are independent
– Subcarriers with high attenuation for user 1 may have good
gain for user 2

29
Channel Dependent Scheduling

Time (ms)

30
P
Proportional
ti lFFair
i Scheduler
S h d l

Schedule the user

with the highest
ratio:

Rk = possible rate of Ch
hannel gain
user k
Tk = average
throughput of user k
ti
time
 Adaptive coding and modulation

QPSK 16QAM and 64QAM modulation (m=2,

• WiMAX and LTE use QPSK, =2 4 and 6)
– Higher m provides higher data rate
g m is subject
– Higher j to more errors
• WiMAX uses Convolutional coding and Turbo coding
– Code rate r between 1/2 and 5/6
• TE use Turbo coding only for user data
– Code rate r between 1/3 and 1

• Lower m and higher r is effective against fading, and vice versa

• But lower m and higher r reduces data rate to and from the user
• WiMAX and LTE select the values of m and r to match the fading channel for each user

32
Modulation methods

BPSK Q_out
b0 Q_out
0010 0110 1110 1010
3
16-QAM
I_out b0b1b2b3
-1 +1
0011 0111 1111 1011
1

QPSK I_out
b0b1 -3 -1 1 3
Q_out
01 11 -1
1
1 0001 0101 1101 1001

I_out
-1 1
0000 0100 1100 1000
-3
3
-1
00 10

33
Modulation methods
64-QAM
b0b1b2b3 b4b5 Q out
Q_out

000 100 001100 011 100 010 100 110 100 111 100 101 100 100 100
7

000 101 001101 011 101 010 101 110 101 111101 101101 100 101
5

000 111 001111 011 111 010 111 110 111 111 111 101 111 100 111
3

000 110 001110 011 110 010 110 110 110 111 110 101 110 100 110
1
I_out
-7 -5 -3 -1 1 3 5 7
-1
000 010 001010 011 010 010 010 110 010 111010 101010 100 010

-3
000 011 001011 011 011 010 011 110 011 111011 101011 100 011

-5
000 001 001001 011 001 010 001 110 001 111 001 101 001 100 001

-7
000 000 001000 011 000 010 000 110 000 111000 101000 100 000

34
 Adaptive coding and modulation

• Note that each user has a different channel condition between the base station and the mobile station

35
 MIMO History

• RX diversity - HF, terrestrial microwave, cellular….

• TX frequency offset diversity & simulcasting for paging - 70’s
• Adaptive array processing in military systems
• TX diversity - 80’s
– frequency offset (channel decoding combining)
– delay (equalizer combining)
• Optimum combining for cellular (multipath channels) - 80’s
• Space-division multiple access - 80’s & 90’s
– angle-of-arrival
g based
– multi-path based (supports co-location & multi-channels per user)
• MIMO - 80’s & 90’s
– Multiple spatial channels using adaptive antenna arrays
– BLAST - successive interference cancellation combined with coding
– Space-Time coding
 MIMO

• Si
Signall ttransmitted
itt d from
f multiple
lti l antennas
t
(Multiple In)
g received byy multiple
• Signal p antennas ((Multiple
p
Out)

• Receiver combines the received signals and optimally combine energy from MxN
channels
• Two main types of MIMO
 Transmit Diversity (also called Alamouti)
 Spatial
p Multiplexing
p g

37
MIMO Modes
 MIMO Modes
• Transmit diversity:
– Same modulation symbols sent from all Tx M antennas
ece e co
– Receiver combines
b es thee ssignal
g a from
o N antennas
a e as
– Useful to increase performance against fading
• Spatial multiple
multiplexing:
ing:
– Different modulation symbols sent from M Tx antennas
– Receiver received the signal from N antennas
– Useful to increase data rate if channel is good
• WiMAX uses up to 2x2. LTE uses up to 4x4

39
Conventional Receiver Diversity


S o h1  h2
2 2

Combining two
channels
h l strength
t th

R
Receiver
i pays the
th costt off antenna
t diversity
di it

40
MIMO 2X1, Transmit Diversityy (Alamouti Codes)

• Example M=2 and N=1 TX Diversity

p Two antennas. Time: Two intervals
• Space:
• Cost moved to transmitter (Base Station)

41
 MIMO 2X2,
2X2 Transmit Diversity

• Take M
M=22 and N N=22
• Diversity order 4

42
 MIMO 2x2,
2x2 Spatial Multiplexing

ro  s o g o  s 1 g 1
r1  s o g 2  s 1 g 3

• Purpose is to increase data rate (2x2 gives twice data rate)

• The 4 gains must be known at receiver
• Simplest way at receiver, matrix inversion:

 ro   g o g 1   s o   sˆo  1  ro 
r   g  
 sˆ   G G  G
H H
 s  r 
 1  2 g
3  1  1  1
G

43
 Turbo Codes
• Turbo codes were proposed by Berrou and Glavieux in the 1993 International Conference in Communications
(ICC ’93)
• Break Through performance, much better than conventional methods
• Features
eatu es of
o turbo
tu bo codes:
– Parallel encoding
– Each encoder is a Systematic encoder
– Interleaving among the encoders
– Iterative decoding

44
 Turbo Encoder

• Source bits are encoded by first encoder

• Source bits are interleaved in a pseudo-random fashion and encoded by second encoder
• Original source bits also transmitted (systematic)
• Overall rate is r=1/3

45
 Turbo Encoder …

LTE Turbo Encoder

46
 Turbo Iterative Decoding

• Decoder 1 uses: original bits, parity 1 & extrinsic 2

• Decoder 2 uses: original bits, parity 2 & extrinsic 1
• Decoder 1 provides extrinsic 1 to improve confidence level to input of decoder 2
• Decoder 2 works now better, and feeds back extrinsic 2 to improve confidence level to input of decoder 1
• Decoder 1 repeats with the better input, provided to decoder 2
• Decoder 2 repeats with the better input, feedback to decoder 1
• . . . . . Many iterations

47
Turbo Coding Performance with iterations

48
 WiMAX vs LTE parameters

Parameter WiMAX LTE

Duplex method TDD FDD and TDD

Bandwidth 5 and 10 MHz 1.25, 3, 5, 10, 15 & 20 MHz

Frame size 5 ms 10 ms with 10 sub-frames

Multiplex Access DL OFDMA OFDMA

Multiplex Access UL OFDMA SC-FDMA

Scheduling speed Every frame (5 ms) Every sub-frame (1 ms)

Subcarrier spacing
p g 10.9 kHz 15 kHz

Maximum DL Data rate (SISO) 46 Mbps (10 MHz band) 50 Mbps (10 MHz band)

Modulation QPSK, 16QAM, 64 QAM QPSK, 16QAM, 64 QAM

Coding for user data Convolution or Turbo Turbo

Diversity MIMO up to 2x2 MIMO up to 4x4

TD & SM TD & SM

49
 Conclusions

• WiMAX and LTE employ similar technologies

• Both will achieve very high data rates
• Both will provide new services
• B h use: OFDMA
Both OFDMA, MIMO
MIMO, TURBO
• LTE has the advantage of large GSM/UMTS customer base
• WiMAX has the advantage of being already in service in few places in USA

50
S
System AAspects off
LTE Ad
Advancedd
Evolution of Radio Access Technologies

802.16m

802.16d/e

• LTE (3.9G) :
3GPP release
l 8~989
• LTE-Advanced :
3GPP release 10+
LTE Release 8 Key Features

• Hi
Highh spectral
t l efficiency
ffi i
– OFDM in Downlink
– Single‐Carrier FDMA in Uplink
• Very low latency
– Short setupp time & Short transfer delayy
– Short hand over latency and interruption time
• Support of variable bandwidth
– 1.4, 3, 5, 10, 15 and 20 MHz
• Compatibility and interworking with earlier 3GPP Releases
• FDD and TDD within a single radio access technology
• Efficient Multicast/Broadcast
 Evolution of LTE-Advanced

• AAsymmetric
t i transmission
t i i bandwidth
b d idth
• Layered
y OFDMA
• Advanced Multi-cell Transmission/Reception Techniques
• Enhanced Multi-antenna Transmission Techniques
• Support of Larger Bandwidth in LTE-Advanced
LTE Advanced
 Asymmetric transmission bandwidth

• Symmetric transmission
– voice transmission : UE to UE
• Asymmetric transmission
– streaming video : the server to the UE (the downlink)
 Layered OFDMA

• The bandwidth of basic frequency block is,

is 15–20 MHz
• Layered OFDMA radio access scheme in LTE-A will have layered transmission bandwidth, support of layered
environments and control signal formats
Advanced Multi-cell
Multi cell Transmission/Reception Techniques

• In LTE
LTE-AA, the advanced multi
multi-cell
cell transmission/reception processes helps in increasing frequency efficiency and
cell edge user throughput
– Estimation unit
– Calculation unit
– Determination unit
– Feedback unit
 Enhanced Multi-antenna Transmission Techniques

• In LTE
LTE-AA, the MIMO scheme has to be further improved in the area of spectrum efficiency,
efficiency average cell through put and cell edge
performances
• In LTE-A the antenna configurations of 8x8 in DL and 4x4 in UL are planned
 Enhanced Techniques to Extend Coverage Area

• Remote Radio Requirements (RREs) using optical fiber should be used in LTE-A as effective technique to extend cell coverage
 Support of Larger Bandwidth in LTE-Advanced

• Peak data rates up to 1Gbps are expected from bandwidths of 100MHz.

100MHz OFDM adds additional sub-carrier
sub carrier to
increase bandwidth
LTE vs. LTE-Advanced
Outlook on candidate technologies

– Novel MU-MIMO algorithms (PHY, MAC)

• Adaptive switching between single-/multi-user/multi-site modes
• Combination of spatial multiplexing and beamforming
– Network MIMO concepts and algorithms (PHY, MAC) for FDD/TDD
• Coherent/non-coherent solutions
• Centralized (e.g.RRH) and distributed (collaboration among Node Bs) solutions
– Dynamic ICIC concepts
• Dynamic exchanges
h off resource bl
blocks
k ultiziation
l among Node d Bs
• Beam Coordination between cells in collaboration
– Schedulers for exploitation of the advanced MIMO and multimulti-site
site features
• Cross-layer optimal resource allocation with advanced MU-MIMO/IFCO features
• Multi-site scheduler with exploitation of the multi-site features
• Interworking and optimization between UL/DL scheduler
 MIMO for LTE ad
advanced
anced FDD

• Overall MIMO recommendations for LTE advanced (FDD):

– Place greatest emphasis on MU-MIMO, since it has the most attractive performance-complexity tradeoff
– SU-MIMO should be pursued to deliver high peak user rates for IMT-Adv requirements
– Increase
I off DL cellll edge
d rates by
b
• Multi-site Collaborative MIMO (constructive data instead of interference)
• Complemented by a combination
– Spatial Interference Coordination (beam coordination)
– Fractional frequency/time reuse Interference Coordination
– Further gains in spectral efficiency are desired on uplink,
• network MIMO with coherently coordinated bases.
MIMO Configurations

MIMO

Single base Multiple

p bases
(Network MIMO)

Co-located Distributed Noncoherent Coherent

antennas antennas (Magnitude only) (Magnitude/phase)
SU-MIMO, Macroscopic Collaborative Coherent
MU-MIMO
MU MIMO MIMO MIMO Network
MIMO
 Single site MIMO eevolutions
Single-site olutions (FDD)

• DL MU-MIMO based on one or both of the following approaches depending on antenna configuration,
configuration cell size and mobility:
– Fixed-beams (e.g., grid-of-beams approach): Suitable for high mobility, can operate without dedicated pilots (but would
benefit from them), works best with closely spaced antennas. >= 4*x xpol. Tx-antennas or 1*4 Tx antennas
– User-specific beams (e.g., ZF): Suitable for low mobility, requires dedicated pilots, but potentially better interference
suppression. 2*2 Tx antennas
• With closelyy spaced
p antennas the same ggiven beam could be applied
pp over the whole bandwidth,, reducingg uplink
p feedback
requirements.
• Techniques (e.g., hierarchical feedback) to reduce CSI feedback requirements.
– MU-MIMO with user-specific beams should be revisited with the target of reduced feedback bandwidth.
• UL MU-MIMO
 Performance improvement with more than 1 transmit antennas at UE (2-4)
– ensure that signaling supports co-channel transmission by multiple users.
Evolved MIMO for IMT-Advanced:
IMT Advanced: Extended Precoding

Combinations of Beamforming and Diversity Transmission

• Beamforming for Multi-User Transmission (SDMA), based on closely spaced antenna elements (0.5
lambda)
• Diversity for link enhancement and/or spatial multiplexing
multiplexing, based on cross-polarized
cross polarized antenna elements
– Requires appropriately optimized codebooks for the antenna weights
– For up to 8 antenna elements in a 4x2 X-pol. configuration ( compact housing)

d t stream
data t 1/2
MS 1
Base- MIMO channel
station
data stream 3 MS 2
Multiuser MIMO and schedulingg for limited feedback

Relative Sum throughput gain

• In Multiuser (MU) MIMO, multiple streams can
be allocated among different users. Mobile speed (kmph) 3 50

• MU eigenmode transmission (MET) uses channel Unitary beamforming

k
knowledge
l dg att th
the T
Tx tto fform non-interfering
i t f i g (baseline) 1.0
.0 1.0
user-specific beams. MET with
hierarchical feedback 1.4 1.28
- Design codebooks whose codewords are indexed
using
i g uplink
li k ffeedback
db k bit
bits. K = 20 users per sector, 1 rx ant per user,
- Aggregate B feedback bits per signaling interval B = 4, M = 4 tx ants (10spacing) Note:
Baseline values normalized to 1 for different velocities
for hierarchical feedback.
11 Users
U estimate
ti t channel
h l and
d ffeedback
db k
MET block diagram 3 A quantized state.
B 22 Base selects users to serve and
User calculates beam weights to
User Beam
Beam-
data maximize sum rate while addressing
streams selection 2 forming
C fairness.
3
3 Data is transmitted.
D
Channel state feedback 1
 Multiple-site
l l MIMO (FDD)
( )
 For UL, application of Network MIMO coherently coordinates a reasonable number of base
stations in Rx
• Standardization issues: pilot structure, signaling and X2 Interface
• Issues like backhaul bandwidth and architecture, channel estimation overhead should be
investigated
 Coherent network MIMO for DL onlyy in TDD mode.
• Feedback requirements in FDD are likely to be too high.
 Possible FDD Solution for DL as a adaptive combination of:
• Multiple Site Non-coherent Collaborative MIMO to leverage Cell edge rate
• Single Site MU-MIMO to leverage the cell average rate
• Single
Si l site
it SU
SU-MIMO
MIMO ((+ TTx-Div)
Di ) tto leverage
l th
the user rate
t
Collaborative/Network MIMO overview

Coordinate transmission and

reception of signals among
multiple bases.

Reduces intercell interference

and improves cell-edge
performance and overall
throughput.

Collaborative MIMO: share user

data and long-term noncoherent
channel information.

Coherent network MIMO: share

user data and short-term
coherent channel information.
Multi-Mode Adaptive
p MIMO for DL/UL
/

Use adaptive MIMO to accommodate demand of MAC layer

higher data rate and wider coverage in next
generation broadband wireless access
• SU MIMO for peak user data rate improvement Cross-layer
design
• MU MIMO for average data rate enhancement
• Collaborative/Network MIMO for cell edge user
data rate boost

election
SU-MIMO
A
uniform

adaptive se
MIMO MU-MIMO
platform
Collaborative/
Network
MIMO
Key technologies in Multi-mode
Multi mode Adaptive MIMO

Serving eNB/ Multi-dimension adaptation

per User Cellular system •Adaptation strategy
•Multi-variable channel measurement
•Low-rate feedback mechanism

SU-MIMO

SU MIMO enhancement
SU-MIMO
•Closed-loop MIMO
Multicast •Iterative MIMO receiver
Anchor

Collaborative/Network
MIMO MU-MIMO
Collaborative/Network
MIMO/Beam Coordination
•Implementation of multi-BS MU-MIMO optimization
collaboration with channel •MU precoding algorithm
information •Trade-off design of scheduler between
eNBs have to be synchronized !!! complexity and performance
 Coherent Network MIMO for UL

• What is it: Interference reduction via coherent receiver coordination between multiple
p bases.
• How does it work: Coordinating base stations compute beamforming weights that maximize SINR (MMSE) for each
user.
• Potential
P t ti l performance
f gains
i off Network
N t k MIMO for f S-sector
t coordination
di ti
Baseline: single-sector (1Tx 4Rx)

What’s needed to make it happen:

 Short-term coherent channel knowledge and user data shared among coordinating bases.
 Backhaul traffic increases by factor S (if all users are In collaboration)
 10% channel knowledge, 90% user data .
 Time and phase-synchronized transmission among coordinating BSs.
Efficient Channel Quality Feedback for IMT
IMT-Advanced
Advanced

• UL feedback channel is a bottleneck for the system

y performance
p in an FDD system.
y A more efficient feedback scheme
provides
– lower resource usage in the uplink and/or
– higher downlink performance through finer granularity of the channel state information knowledge at the base
station
• Compression / sourcecoding of channel state information feedback
– based, e.g., on Wavelets (or other transformations)
• allows variable frequency resolution over the bandwidth
• e.g.,
g UE adaptively
p y pprovides highg resolution of CQI
Q on ggood subcarriers / resource blocks, & low resolution
on bad resource blocks
• Hierarchical Feedback approach
• successive refinement of the quantization with imperfect channel state information at the Tx.
Tx
 Conclusions

 Mix of these technologies allows to meet the IMT adv performance requirements
 The introduction/improvement of MU-MIMO in DL and UL has a high potential to
boost the cell average
g rate
 Co-MIMO for DL can be applied to FDD system to improve cell edge performance
and average cell capacity
• About 70% improvement for cell edge rate rate compared with SU-MIMO
• 25% improvement in average sector capacity compared with SU-MIMO
– Network MIMO can be applied for the UL (FDD) for the DL/UL(TDD)
• 25% improvement for cell average rate compared to MU-MIMO (further
improvements from single site MU-MIMO)
• Factor 3.4 gain for cell edge rate
 S lf Organizing
Self O ii &
Optimizing Networks (SON)
Target: Simplified Net
Network
ork Operation

• Self Organizing Network (SON) technologies

Self-Organizing-Network
– 100% Plug and Play
– Fully decentral. OMC less Network Management (prio for pico/femto layer)
• Self-protection against malicious resource usage (multi-vendor problem)
• Multi-RAT operation (intra 3GPP and inter 3GPP)
– Self-configuration / optimization for heterogeneous networks
(3GPP / non
(3G o 3G
3GPP))
– Generic protocols and measurements
• Generic parameters for
– Handover decisions
– Load balancing
– QoS optimization
• Multi-operator
Multi operator networks
• RAN sharing
• Equipment sharing
Evolution: Phased Approach for Self-x
Self x (SON) introduction

– First step (LTE-R8):

• focus mainly on configuration use cases needed for first deployments
• NEM centric automated configuration and tool based optimisation
• Self-x
lf support functions
f decentralized
d l d in eNB (for
(f configuration
f andd optimisation use cases))
• tight control and surveillance in OMC

– Second step (beyond LTE R8):

• decentralised “NEM less” architecture (Pico & Femto Layer)
• Complete Self
Self-xx functions put to eNB
• NM/OSS: performance and alarm management,
• NM/OSS: control/tuning of Self-x use cases requiring
– deeper system performance analysis and simulation,
– further standardisation required
Evolution: Phased Approach for Self-x (SON) introduction
• Release 8:
– RAN configuration use cases:
– Add/Remove cell incl. power saving cell (Auto download of initial radio parameters from OMC)
– Neighbourhood relation configuration and optimisation for LTE

• Release 9 and +: tools for RAN deployment new

– RAN optimization use cases planning,
configuration
site,
add new cell,
failure cases
performance
optimisation
capacity upgrade
– Cellll outage compensation and
d optimisation
ti i ti

– LTE handover parameter optimisation

– Interference optimisation for LTE
– Load
L d balancing
b l i forf LTE
conventional
– QoS optimization use cases parameter
configuration
self-configuration self-optimisation

– Scheduler operation optimisation for LTE

– MIMO Mode Selection Optimisation for LTE
Use Case “LTE
LTE Handover Parameter Optimisation
Optimisation”

• Self-optimisation of initially configured HO parameters

– Optimisation goals
Signal
• Minimisation of HO failure rates for intra-LTE strength Source cell

• Avoidance of ping-pong effects HH

• Enhancement to Multi-RAT HO HA
Neighbor cell
– Optimisation
O i i i approach: h SSelf-optimisation
lf i i i off
HO parameters leading to UE handover request TTTA

• HO thresholds,, hysteresis,
y , Cell Individual Offset TTTH
(CIO),time to Trigger (TTT) after analysis of
handover
Hand-
Addition
Addi i Ti
Time
– Challenge: user throughput at HO (cell edge) Event over
Event

• Considering QoS at cell edge during handover as

constraints
Use Case “Interference
Interference Coordination in UL and DL
DL”
• Dynamic or semi-static interference coordination of radio
resources (example: frequency case) resource grant (Tx pwr on
certain frequency subsets)

– Possible optimisation goals resource

request

• Cell edge bitrate, improved fairness, load balancing,

increased number of real time users, network capacity eNB #2

– Power restriction scheme and power attenuation eNB #1

• Indication of upper limit of Tx power per PRB

eNB #3
relative to the rated output power
• Exchange of upper limits of the Tx power per PRB
and resource restrictions between neighbour eNBs
• over X2 interface in intervals of 200 ms to 1 s
Use Case “Scheduler
Scheduler optimisation
optimisation”

• Optimisation goals
– self-optimisation of user -, cell-, cell edge throughput & delay according to operator
ppreferences with weightings
g g and fairness pparameter
– self-optimisation of network service availability per QoS label
• Optimisation
p approach
pp for Q
QoS and scheduler configuration
g pparameters
– indication of estimated impact on performance and resulting QoS based on target derived
from Off-line System Simulations
– adaption of scheduler operation to actual traffic mix
– PFMR (Proportional Fair with Minimum Rates) scheduling for tuning of cell edge bit rate
andd cost versus fairness
f proportionall to experiencedd radio
d conditions
d
Use Case “MIMO
MIMO Mode Selection Optimisation for LTE”
LTE

• Optimisation
p ggoal of MIMO modes switchingg
• Optimum service provisioning among attached UEs
• Cell edgeg data rate and total cell throughput
gp
• Optimisation of network due to insufficient radio condition (SINR) at cell
edge, and service availability per QoS label
• Optimisation approach
– Evaluation of mapping of link characteristic (rank, SINR) to MIMO modes
– Configuration of MIMO thresholds and MIMO-mode switching criterions
(diversity, beamforming, spatial multiplexing for SU MIMO and MU MIMO)
supported
t d byb targets
t t derived
d i d from
f Off-line
Off li SSystem
t SiSimulation
l ti
Vision of fully decentralised self
self-optimisation
optimisation
NM OSS
– Network management in NM OSS
Itf-N
• network planning Network Management X2-Itf

• alarm and performance monitoring

• high
hi h llevell performance
f ttuning
i
• open Itf-N  performance
– “NEM less” network management  high level network monitoring
performance tuning
p g  KPIs
– Fully
F ll autonomous,
t di
distributed
t ib t d RAN optimisation
ti i ti  alarms
– Self-x functions in UE and eNB
• measurements, UE location info LTE RAN
• alarms, status reports, KPIs self-x
• distributed self-x algorithms
– Self-x information exchange eNB
via X2 self-x
self-x

– Multi-vendor interoperability RAN self-

optimization
supported via X2 (to support Pico & Femto
p y
deployments )
eNB
NB
eNB
 Conclusions

• Significantly improved radio network management by SON:

– emphasis on performance tuning and supervision
– 100% plug and play
– continuous,
ti automated
t t d radio di network
t k optimisation
ti i ti
w.r.t. operator preferences
– innovative
i i techniques
hi ffor performance
f optimisation
i i i
(scheduler, MIMO modes)
– considerable
d bl effort
ff reduction
d for
f operators
Y Questions,
Your Q i Wi Wishes,
h PProposals?
l?
Th k YYou for
Thank f Your
Y Attention!
A i !
B k
Backup
LTE Architecture

• SAE architecture
hit t [3GPP TS 23.401]
23 401]

GERAN
Gb HSS PCRF

UTRAN Iu GPRS Core S6 Rx+

S7
X1 S3 S4

Operator IP
S1
S SGi services
eNB
NB SAE PDN
MME UPE (including IMS,
S11 GW SAE GW
S5 PSS, ...)
X1 X2 aGW
Evolved Packet Core S2

eNB
Non-3GPP IP
Access
Evolved RAN

88
 Functions of eNB

• Terminates RRC,
RRC RLC and MAC protocols and takes care of Radio Resource Management functions
– Controls radio bearers
– Controls radio admissions
– Controls
C l mobility
bili connections
i
– Allocates radio resources dynamically (scheduling)
– Receives measurement reports from UE
• Selects MME at UE attachment
• Schedules and transmits paging messages coming from MME
• Schedules and transmits broadcast information coming from MME & O&M
• Decides measurement report configuration for mobility and scheduling
• Does IP header compression and encryption of user data streams
Functions of aGW I

• Takes care of Mobility Management Entity (MME) functions

– Manages and stores UE context
– Generates temporary identities and allocates them to UEs
– Checks
Ch k authorization
h i i
– Distributes paging messages to eNBs
– Takes care of security protocol
– Controls idle state mobility
– Control SAE bearers
– Ciphers & integrity protects NAS signaling
Functions of aGW II

• Takes care of User Plane Entity (UPE) functions

– Terminates for idle state UEs the downlink data path and triggers/initiates paging when downlink data
arrive for the UE.
– Manages and stores UE contexts,
contexts e.g.
e g parameters of the IP bearer service or network internal routing
information.
– Switches user plane for UE mobility
– Terminates user plane packets for paging reasons
LTE User Plane
l

UE eNB aGW

IP IP

PDCP PDCP

RLC RLC

MAC MAC

PHY PHY

S1
Protocols: Radio interface
Protocols: eNB - aGW

S1
 Generic Frame Structure

• Allocation of physical resource blocks (PRBs) is handled by a scheduling function at the 3GPP base station
(eNodeB)

Frame 0 and frame 5 (always downlink)

 LTE Frame Structure

• Frame is 10 ms,
ms divided into 10 sub
sub-frames
frames

96
Resource Grid

• One frame is 10ms

 10 subframes

• One subframe is 1ms

 2 slots
• One slot is 0.5ms
 N resource blocks
[ 6 < N < 110]]
• One resource block is 0.5ms and contains 12
subcarriers from each OFDM symbol
LTE spectrum (bandwidth and duplex) flexibility
 LTE Downlink Channels

Paging Control Channel

Paging Channel

Physical Downlink Shared Channel

LTE Uplink Channels

Random Access Channel

CQI report
Physical Uplink Shared Channel

Physical Radio Access Channel

 WiMAX Frame Structure

• Frame duration is 5 ms

101
WiMAX User Data Tx

Transmitter Baseband Processing

102
WiMAX User Data Rx

Receiver Baseband Processing

103
LTE User Data Tx

104
LTE User Data Rx

105