LTEの実現とさらなる飛躍に向けて

LTE: Today's Implement and

Tomorrow's Advancement

NEC
Motoya Iwasaki

© NEC Corporation 2009

 Contents
Introduction
LTE Architecture
Physical Channel Structure and Procedure
Signal Processing
Algorithm (eNB)
Architecture
Radio Resource Management (eNB)
LTE Advanced
Conclusion

Ｐage 1 © NEC Corporation 2009

 16
16 16 16 Introduction
14 16 14
14 14
Trend for Mobile
16
12 14 12 Wireless Broadband
12 12
14
10 12 10
10
Two108Streams
12
10 for Mobile GSM GSM GSM
GSM
8 Broadband
8 8
10 HSPA GSM HSPA HSP
6 8 16
HSPA
6 Throughput
6 6 LTE GSMHSPA
ITU Family 8
6
[Mbps]14 Fixed Mobile
LTE
LTE LTE
4 4
16 12
Convergence
4 16 LTE
HSPA
4
4HSPA, LTE etc.
14
 W-CDMA,2
6
12
10 14

12 802.16e LTE
4G
GSM
2 2
8
802.11n 802.16m
 CDMA2000, 4 EV-DO, UMB etc.
16
2 10 6
10 HSPA

0 2 14 802.16d 8
LTE GSM LTE
GSM

0 1000 8 HSPA

0 0 0 12
4 6 802.11a/g HSPA LTE
2 00.2 0 0.20.4 0.2 0.40.6 0.4 0.8 1 1.2
16
-2 0 6
0.62
0.8
4
1 1.2
IEEE Family
0 -2 0.2-2 0.4 0.6 10414 0.8 0.6 1 0.8 2
1.2 1 1.2
HSPA+
LTE

-2 100 0 GSM
0 0
-2
0.2 0.4 0.6
8 16 0.80.2 0 1 0.6
802.11b
1.2 1
 Wi-Fi, WiMAX etc. 12
2 -2
0
6 14
0.4
-2
0 0.2 0.40.8 0.6 0.8 11.2
HSPA 1.2

0 0.2 0.4 0.6

10
0 0.8 1 HSPA 1.2 LTE
-2 4 120 0.2 0.4 0.6 0.8 1 1.2 GSM
-28
102 10 HSPA
6
R99/R4 LTE
GSM
0 8
1 4 HSPA
0 0.2 0.4 0.6 0.8 1 1.2
-2 6 GSM/GPRS/EDGE
2 LTE
4
0 Low
-2
2 0 0.2 0.4 0.6 0.8 1 1.2
Mobility
0
High
0 0.2 0.4 0.6 0.8 1 1.2
-2
Realization of Mobile Broadband Services

Ｐage 2 © NEC Corporation 2009

 LTE Architecture (SAE)
One RAN node : eNB
MME
Core Network Operator’s IP Evolved Packet Core (SAE) logical node
service network NAS Security

SGSN GGSN Serving Gateway MME Idle State Mobility

SAE Bearer Control

IP SAE Gateway
logical node
S1-MME Mobility Anchoring
S1-MME
Iu Iu
eNB
logical node
Iur S1-U Inter Cell RRM
RNC S1-U Radio Admission
Control
RNC Radio Bearers
ATM/IP IP Control
Iub transport transport Handover
Iub
NB NB Scheduling
eNB
RRC
eNB X2
PDCP
MAC/RLC
UTRAN <HSPA+> eUTRAN (LTE) PHY

MME: Mobility Management Entity  Functions in 3G RNC are moved to eNB in LTE
S-GW: Serving Gateway  Call Processing related functions are moved to eNB.
EPC: Evolved Packet Core
 C-plane latency and U-plane latency can be minimized due to
eNB: evolved NodeB reduction of layer of network elements and optimization of
signalling.
Ｐage 3 © NEC Corporation 2009
 Physical Channel Structure
<Downlink> DL PhCH Structure
PDSCH PCFICH PHICH
Transmit Data f Reference Signal
PDCCH
Assign PDSCH/PUSCH

(Control Signal)
SCH and BCH is
PHICH PDSCH (Shared Data)

System BW
transmitted at certain
Indicate HARQ-ACK for UL

PDCCH
subframe only
PCFICH BCH
Indicate PDCCH symbol
BCH
Transmit Broadcast S-SS P-SS
channel
Synchronization Signal
For UE synchronization
t
PRACH preamble and SRS is
f UL PhCH Structure transmitted at pre-assigned
<Uplink> subframe only
PUSCH PUCCH (Control Signal)
Transmit Data
PUCCH
System BW

Transmit ACK/NACK, CQI,

SRS
SR PUSCH (Shared Data)
PRACH
Transmit Random Access
Preamble
SRS RACH Preamble
For UL CQI measurement PUCCH (Control Signal)
1slot = 0.5 ms
1sub-frame= 1.0 ms
Ｐage 4 © NEC Corporation 2009
 Physical Channel Procedure
Synchronization Procedure
eNB
DL
S-SS P-SS DL RS PBCH PDSCH

UL
UE
Cell-Id Id 0～2 Cell Specific System System
Group RS Information Information
0～167 (MIB) (SIB)

Cell-Id Detect Broadcast Information

0～503
Random Access Procedure
RRC Connected
Time
Alignment PDCCH PDSCH
eNB
DL Data Data
Schedule PUSCH
UL Info Data Data
RACH
UE
Message 1 Message 2 Message 3 Message 4 Data Transfer
Random Access Random Access RRC Connection RRC Connection
RRC: Radio Resource Control Preamble Response Request Setup

Ｐage 5 © NEC Corporation 2009

 Physical Channel Procedure
DL Data Transmission Procedure
Scheduling HARQ
Link adaptation Scheduling
Link adaptation
PDCCH PDSCH
eNB
RS
DL Data Data
PUCCH or PUCCH or
PUSCH Scheduling
PUSCH
UL Info.
CQI CQI ACK
UE
CQI
HARQ
Measure

UL Data Transmission Procedure

HARQ
CQI Scheduling Scheduling
Measure Link adaptation Link adaptation
PDCCH PHICH
eNB
DL ACK
PUSCH
PUCCH Scheduling
Grant Data Data
UL SR
BSR
UE
SRS (Buffer Status Report)
(Sounding RS) (Scheduling Request) HARQ

Ｐage 6 © NEC Corporation 2009

 Physical Channel Procedure
- Resource Assignment
Resource Allocation
PDCCH assigns PDSCH of the same subframe
PDCCH assigns PUSCH of 4 subframe later
Dynamic scheduling
Resource is allocated by PDCCH subframe by subframe
Semi-persistent scheduling
Period is pre-assigned by RRC
Initial transmission is assigned by PDCCH
Transmission continues periodically.
f PDCCH PUSCH
PDSCH
DL UE1 UE1 UE1
Info. UE2
UE2 UE4
UL UE1
grant UE2
UE3 UE4 UE5
DL UE3
PCFICH UE2 UE2
DL UE4
UE4 UE5 UE3
UL UE5
grant UE6 UE2 UE7
UE7
DL UE5 UE3 UE6 t

4 subframe
Actually, PDCCH data for each
user is interleaved each other.

Ｐage 7 © NEC Corporation 2009

 Physical Channel Procedure
HARQ
ACK/NAK retransmission

DL HARQ PDSCH PUCCH PDSCH

UE eNB
- CH Decode - CH Decode
- CH Coding - Scheduling
- CH Coding
UL HARQ PDCCH PUSCH PDCCH/HPHCH
UE eNB
- CH Decode - CH Decode
- CH Coding - Scheduling
- CH Coding
TTI Bundling
Single transport block is coded and transmitted in Consecutive subframes.
The same hybrid ARQ process number is used in each of the bundled subframes
No change in HARQ timing relation
Coding RM

RV=0 RV=1 RV=2

ACK/NAK retransmission

HARQ process number #0 #0 #0 #3 #4 #5 #6 #7 unused #0 #0 #0 #3 #4 #5 #6 #7

bundle
HARQ RTT HARQ RTT HARQ RTT

Ｐage 8 © NEC Corporation 2009

 Signal Processing (UL)
PUSCH Demodulation
Channel estimation & Equalization on Frequency domain
Both RS in subframe is used to improve performance.

FFT FFT FFT FFT FFT FFT FFT FFT FFT FFT FFT FFT FFT FFT

Equalization IDFT

RS* RS*
Channel Combine
Estimation etc.
Noise Noise
Reduction Reduction

PUCCH Demodulation
Similar to PUSCH
Format 1/1a/1b Format 2/2a/2b

Data
ZC Despread Despread ZC Despread

RS Despread
CH Estimation CH Estimate
Combine

Ｐage 9 © NEC Corporation 2009

 Signal Processing (UL)
SRS (Sounding RS) Detection
Users are multiplexed by Distributed-FDM and cyclic shift based CDM
Frequency domain CQI is measured

D-FDM RS* User

de-mapping de-MUX
Profile on
FFT IFFT FFT F-domain

User1 User2 User3

User4

PRACH Detection
6bit information is expressed by Cyclic shift and Root sequence
RTD and Cyclic
RS*
shift is detected
Decimation & Peak
Down-sample FFT IFFT Detect

CP RACH Preamble
Peak for frequency offset of +fo
Cyclic shift 0 High Speed Case Peak for no frequency offset
Cyclic shift 1 Peak for frequency offset of -fo
Cyclic shift 2

Cyclic shift N-1

Ｐage 10 © NEC Corporation 2009

 Signal Processing (UL)
MIMO Receiver
MLD : For DL only (Can not be applied to DFT-spread-OFDM (SC-FDMA))
MMSE : Rather poor performance
Turbo-SIC : 2~3 heavier processing load
PUSCH PDCCH
4 subframe

FFT FFT FFT

SIMO MMSE Equalization Equalization Equalization

Channel
Estimation IDFT IDFT IDFT
Decode (Turbo) Decode (Turbo) Decode (Turbo)
MAC Scheduler MAC Scheduler MAC Scheduler
PDCCH PDCCH PDCCH
Gen Gen Gen

MIMO MMSE Equalization

Equalization Channel
Channel Estimation
Estimation IDFT IDFT
Decode Decode Decoder
FFT Equalization IDFT
(Turbo) (Turbo) (Turbo)

MIMO Turbo-SIC Equalization

Channel
Channel
IDFT Estimation
Estimation
Decode (Turbo) Decoder
FFT Equalization IDFT
(Turbo)
Encode

DFT
DFT Encoder

Ｐage 11 © NEC Corporation 2009

 Signal Processing (DL)
DL Demodulation
OFDM demodulation based on RS signals
PDSCH Resource Allocation
Type 0: RB groups are allocated by Bit Map.
Type 1: RBs in selected 'Distributed" RBG subset are allocated by Bit Map.
Type 2: Start position and Width is assigned.
Localized type and Distributed type
Type 0
RBs are grouped
to RBG.

RBGs are selected.

0 1 1 0 0 1 0 0 1 1 1 0 1 0 0 1 1 1 0 0 0 0 0 0 0
Type 1
RBs are grouped
to RBG subset.
First, one of RBG
subset is selected.
Then, RBs in RBG
subset are selected.
1010 1100 0010 0101 0000 0011

Type 2
Start RB and width is assigned.

Localized type
30
24
Distributed type
Slot 0
Slot 1

Ｐage 12 © NEC Corporation 2009

 Signal Processing
Hardware Architecture (Implementation)
WCDMA Chip Rate Symbol Rate
Path CH
Search Est

De Rate Turbo
RAKE
spread Match etc. Decode

Initial ASIC (H/W) DSP (F/W) FPGA

ASIC (H/W) FPGA

Current
DSP (ASSP + F/W)

LTE CH
Est

Equa Rate Turbo

FFT IDFT
lizer Match etc. Decode

Ideal H/W DSP or FPGA H/W FPGA H/W

Not necessary
for DL (UE)

Ｐage 13 © NEC Corporation 2009

 Signal Processing
Hardware Architecture Trend (eNB)
DSP w/ ASSP Others
• Flexible & High performance
DSP DSP DSP
Core Core Core Processor
Address Control
Other genrator Unit
FFT DFT Turbo address context pointer
ASSP
Memory bank memory interface
Sub memory

FPGA bank

Sub memory
PN PN PN PN
bank

Sub memory

External bus
bank

Sub memory
PN PN PN PN
CH Equa Rate CPU bank

Sub memory
Est lizer Match etc. bank

Sub memory
bank PN PN PN PN
FFT DFT Turbo Other Sub memory
bank

Sub memory PN PN PN PN
bank

NEC Laboratories
eNB equipment example
20MHz x 3sector

Ｐage 14 © NEC Corporation 2009

 RRM (Radio Resource Management)
Scheduling
10 User#1
Trigger for UL Scheduling User#2
8
• Buffer Status report (BSR) :PUSCH scheduled User#3

RX Level (dB)
user 6

• Scheduling request (SR) : Synchronous user 4

(PUCCH allocated) 2

• PRACH : Asynchronous users 0

Frequency Selective Scheduling -10 -5 0

Frequency (MHz)
5 10

• DL: Based on CQI (Channel Quality Indicator)

reported by UE User# User#
User# User# User#
• UL: Based on CQI measured by eNB using SR 1 2 3
1 3

Frequency Selective Scheduling

MIMO
• DL: Based on PMI (Pre-coding Matrix Indicator) and RI (Rank
Indicator) reported by UE
• UL: Based on Channel Estimates measured by eNB using SRS
Inter Cell Interference
• FFR
• ICIC

Ｐage 15 © NEC Corporation 2009

 RRM (Radio Resource Management)
FFR (Fractional Frequency Reuse)
To improve Throughput of User at Cell Edge
Edge users is restricted to fraction of transmission bandwidth (i.e. reuse 1/3
for edge users)
Center users can use entire bandwidth (reuse=1)

frequency
１
eNodeB
3 Sector 1
１ 2 １ Sector 2
3 3 Sector 3

2 １ 2
UE
3
Reuse set
2
Reuse set
For all users Reuse set
For users in Sector 1 Reuse set
For users in Sector 2
For users in Sector 3
Ｐage 16 © NEC Corporation 2009
 RRM (Radio Resource Management)
ICIC (Inter-Cell Interference Coordination)
Use X2 Interface (btw eNB) to exchange following information for UL
• HII (High Interference Indicator)
• OI (Overload Indicator)

Notify strong interference

to the adjacent cell Assign that RB to the center user

HII
Observe strong interference Reduce Tx Power
from the adjacent cell

OI
If receive HII, assign that RB to center user to minimize interference.
If receive OI, reduce transmission power of "suspect" user.

Ｐage 17 © NEC Corporation 2009

SON (Self Organizing/Optimizing Network)
eNB power on
Motivation for SON (or cable connected)

Cell size will decrease for future

generation
Radio network control is quite
complicated (Handover, Interference
coordination etc.)
a-1 : configuration of IP address
(A) Basic Setup and detection of OAM
a-2 : authentication of eNB/NW
Self-Configuration
a-3 : association to aGW

(pre-operational state) a-4 : downloading of eNB software

TCO (Total Cost of Ownership) need (and operational parameters)

to reduced
Mechanism to control network (B) Initial Radio b-1 : neighbour list configuration
Configuration
automatically is demanded to reduce b-2 : coverage/capacity related
parameter configuration
CAPEX /OPEX. Self-Optimisation
(operational state)

(C) Optimization / c-1 : neighbour list optimisation

Adaptation
c-2 : coverage and capacity control

Energy Saving,
EPC Self Healing
OMC
Network Parameter
Configuration

Neighbor List,
Handover Parameter UE Coverage Optimization,
Optimization eNB Load Balancing

Ｐage 18 © NEC Corporation 2009

 LTE Advanced
Measure Items now being studied
Bandwidth extension 20MHz~100MHz
Uplink access scheme
MIMO
• DL 8x8, Multi user MIMO
• UL 4x4, Single user MIMO
CoMP (Coordinated Multipoint Transmission/Reception)
• UL (no impact on Uu)/DL, Intra-eNB/Inter-eNB
• Joint-processing/transmission, Coordinated scheduling and/or beamforming
Relay
• (Layer-1 Relay=Repeater), Layer-2,3 Relay
Schedule
2007 2008 2009 2010 2011
LTE CR phase

3GPP RAN #38 #39 #40 #41 #42 #43 #44 #45 #46 #47 #48 #49

WS 2nd WS

Study Item Work Item

ITU-R WP5D
#1 #2 #3 #4 #5 #6 #7 #8 #9 #10 #11 #12

Proposals
Evaluation
Consensus
Specification
Ｐage 19 © NEC Corporation 2009
 LTE Advanced
Bandwidth Extension
Use continuous Component Carriers if continuous frequency available

f
Component Carrier (= LTE Rel-8 ≦ 20MHz)

Use distributed Component Carriers if continuous frequency not available

e.g. 5MHz e.g. 10MHz
@800MHz Band @2GHz Band f
Component Carrier Component Carrier Component Carrier
(= LTE Rel-8 ≦20MHz)

UL Access Scheme
More flexible Scheduling
• Assign only preferred frequency for the UE
PAPR of UE transmission degraded

LTE Rel-8
Contiguous
(SC-FDMA)
LTE-A f
Non-Contiguous
(CL DFT-s-OFDM) f
Component Carrier (～20MHz)
Ｐage 20 © NEC Corporation 2009
 LTE Advanced
UL MIMO Extension
Multi User MIMO : Already covered by LTE (Rel-8)
Single User MIMO
• Spatial bundling of HARQ parameter (ACK, RV etc.)
• Layer shifting in time domain
• Pre-coding
Spatial bundling of HARQ parameter
Layer Shifting Pre-coding
Coding  W00

   

Modulation Layer  

 
  
Coding Mapping 
 

W33 

TX Diversity Modulation

DL MIMO Extension
Pre-coding Code Book
• Extended for 8x8
Multi User MIMO
• Improvement from Rel-8
Downlink RS
Extended for 8 antenna transmission
Cell Specific RS for Measurement and UE Specific RS for demodulation

Ｐage 21 © NEC Corporation 2009

 LTE Advanced
Coordinated Multipoint Transmission/Reception (CoMP)
Intra-eNB -Joint-scheduling

RRH RRH

RRH: Remote RF Head

Centralized eNB Improve Cell Edge

Inter-eNB - Coordinated scheduling
Throughput

eNB eNB

Backhaul (X2)

Relaying
Layer 3 or 2 relay Rural Area
eNB

Relay Blocked Area

Hot spot
Relay
Relay
Ｐage 22 © NEC Corporation 2009
 Conclusion
Signal Processing
Effective Algorithm and Architecture is desired for
• Performance Improvement (Throughput, MIMO scheme)
• Power Consumption Reduction (Of course for UE, also for eNB)
• Cost Reduction
• Flexibility (Standard and Environment is changing)

Radio Resource Management (eNB)

It is key technology to increase cell capacity
• There may be much room for improvement

LTE Advanced
Further Standard improvement is being studied.
 Further improvement for Algorithm and Architecture needed..

Ｐage 23 © NEC Corporation 2009

 Thank you!

Ｐage 24 © NEC Corporation 2009