LTE Evolution: From Release 8 to Release 10

UMTS Long Term Evolution
(LTE)

Reiner Stuhlfauth
Reiner.Stuhlfauth@rohde-schwarz.com

Training Centre
Rohde & Schwarz, Germany

Subject to change – Data without tolerance limits is not binding.
R&S® is a registered trademark of Rohde & Schwarz GmbH & Co. KG. Trade names are trademarks
of the owners.
 2011 ROHDE & SCHWARZ GmbH & Co. KG
Test & Measurement Division
- Training Center -
This folder may be taken outside ROHDE & SCHWARZ facilities.

ROHDE & SCHWARZ GmbH reserves the copy right to all of any part of these course notes.
Permission to produce, publish or copy sections or pages of these notes or to translate them must first
be obtained in writing from
ROHDE & SCHWARZ GmbH & Co. KG, Training Center, Mühldorfstr. 15, 81671 Munich, Germany

Technology evolution path
2005/2006 2007/2008 2009/2010 2011/2012 2013/2014

GSM/ EDGE, 200 kHz EDGEevo VAMOS
DL: 473 kbps DL: 1.9 Mbps Double Speech
GPRS UL: 473 kbps UL: 947 kbps Capacity

UMTS HSDPA, 5 MHz HSPA+, R7 HSPA+, R8 HSPA+, R9 HSPA+, R10
DL: 2.0 Mbps DL: 14.4 Mbps DL: 28.0 Mbps DL: 42.0 Mbps DL: 84 Mbps DL: 84 Mbps
UL: 2.0 Mbps UL: 2.0 Mbps UL: 11.5 Mbps UL: 11.5 Mbps UL: 23 Mbps UL: 23 Mbps

HSPA, 5 MHz
DL: 14.4 Mbps
UL: 5.76 Mbps

LTE (4x4), R8+R9, 20MHz LTE-Advanced R10
DL: 300 Mbps DL: 1 Gbps (low mobility)
UL: 75 Mbps UL: 500 Mbps

1xEV-DO, Rev. 0 1xEV-DO, Rev. A 1xEV-DO, Rev. B
cdma DO-Advanced
1.25 MHz 1.25 MHz 5.0 MHz DL: 32 Mbps and beyond
2000 DL: 2.4 Mbps DL: 3.1 Mbps DL: 14.7 Mbps UL: 12.4 Mbps and beyond
UL: 153 kbps UL: 1.8 Mbps UL: 4.9 Mbps

Fixed WiMAX Mobile WiMAX, 802.16e Advanced Mobile
scalable bandwidth Up to 20 MHz WiMAX, 802.16m
1.25 … 28 MHz DL: 75 Mbps (2x2) DL: up to 1 Gbps (low mobility)
typical up to 15 Mbps UL: 28 Mbps (1x2) UL: up to 100 Mbps

November 2012 | LTE Introduction | 2

3GPP work plan

GERAN
EUTRAN
New
RAN
Phase 1
Phase 2, 2+ UTRAN
Rel. 95
… Rel. 97
Rel.7 Rel. 99 Up from Rel. 8
…
Rel. 7
Rel. 9

Also contained in
Rel. 10

Evolution

Overview 3GPP UMTS evolution

HSDPA/ LTE and LTE-
WCDMA
WCDMA HSPA+
HSUPA HSPA+ advanced

3GPP 3GPP Study
3GPP
release 3GPP Release 99/4 3GPP Release 5/6 3GPP Release 7 3GPP Release 8
Item initiated
Release 10
App. year of 2005/6 (HSDPA)
network rollout 2003/4 2008/2009 2010
2007/8 (HSUPA)

Downlink LTE: 150 Mbps* (peak) 100 Mbps high mobility
384 kbps (typ.)
384 kbps (typ.) 14 Mbps (peak) 28 Mbps (peak) HSPA+: 42 Mbps (peak)
28 Mbps (peak)
data rate 1 Gbps low mobility

Uplink LTE: 75 Mbps (peak)
128 kbps (typ.)
128 kbps (typ.) 5.7 Mbps (peak) 11 Mbps (peak) HSPA+: 11 Mbps (peak)
11 Mbps (peak)
data rate

Round
Trip Time ~ 150 ms < 100 ms < 50 ms LTE: ~10 ms

*based on 2x2 MIMO and 20 MHz operation


Why LTE?
Ensuring Long Term Competitiveness of UMTS
l LTE is the next UMTS evolution step after HSPA and HSPA+.
l LTE is also referred to as
EUTRA(N) = Evolved UMTS Terrestrial Radio Access (Network).
l Main targets of LTE:
l Peak data rates of 100 Mbps (downlink) and 50 Mbps (uplink)
l Scaleable bandwidths up to 20 MHz
l Reduced latency
l Cost efficiency
l Operation in paired (FDD) and unpaired (TDD) spectrum


Peak data rates and real average throughput (UL)
100
58

11,5 15
10
5,76
Data rate in Mbps

5
2 2 1,8
2
0,947
1
0,473 0,7
0,5
0,174 0,153
0,2
0,1 0,1 0,1
0,1

0,03

0,01
GPRS EDGE 1xRTT WCDMA E-EDGE 1xEV-DO 1xEV-DO HSPA HSPA+ LTE 2x2
(Rel. 97) (Rel. 4) (Rel. 99/4) (Rel. 7) Rev. 0 Rev. A (Rel. 5/6) (Rel. 7) (Rel. 8)

Technology

max. peak UL data rate [Mbps] max. avg. UL throughput [Mbps]


Comparison of network latency by technology
800
158 160
710
700
140

600
120

3G / 3.5G / 3.9G latency
2G / 2.5G latency

500
100
85
400 80
320 70

300 60

46
190
200 40

100 20
30
0 0
GPRS EDGE WCDMA HSDPA HSUPA E-EDGE HSPA+ LTE
(Rel. 97) (Rel. 4) (Rel. 99/4) (Rel. 5) (Rel. 6) (Rel. 7) (Rel. 7) (Rel. 8)

Technology

Total UE Air interface Node B Iub RNC Iu + core Internet


Round Trip Time, RTT
•ACK/NACK
generation in RNC MSC
TTI
Iub/Iur Iu
~10msec

Serving SGSN
RNC
Node B

TTI
=1msec
MME/SAE Gateway

eNode B
•ACK/NACK
generation in node B


Major technical challenges in LTE

New radio transmission FDD and
schemes (OFDMA / SC-FDMA) TDD mode

MIMO multiple antenna Throughput / data rate
schemes requirements

Timing requirements Multi-RAT requirements
(1 ms transm.time interval) (GSM/EDGE, UMTS, CDMA)

Scheduling (shared channels, System Architecture
HARQ, adaptive modulation) Evolution (SAE)


Introduction to UMTS LTE: Key parameters

Frequency
UMTS FDD bands and UMTS TDD bands
Range

1.4 MHz 3 MHz 5 MHz 10 MHz 15 MHz 20 MHz
Channel
bandwidth,
1 Resource 6 15 25 50 75 100
Block=180 kHz Resource Resource Resource Resource Resource Resource
Blocks Blocks Blocks Blocks Blocks Blocks

Modulation Downlink: QPSK, 16QAM, 64QAM
Schemes Uplink: QPSK, 16QAM, 64QAM (optional for handset)

Downlink: OFDMA (Orthogonal Frequency Division Multiple Access)
Multiple Access
Uplink: SC-FDMA (Single Carrier Frequency Division Multiple Access)

Downlink: Wide choice of MIMO configuration options for transmit diversity, spatial
MIMO
multiplexing, and cyclic delay diversity (max. 4 antennas at base station and handset)
technology
Uplink: Multi user collaborative MIMO

Downlink: 150 Mbps (UE category 4, 2x2 MIMO, 20 MHz)
Peak Data Rate 300 Mbps (UE category 5, 4x4 MIMO, 20 MHz)
Uplink: 75 Mbps (20 MHz)


LTE/LTE-A Frequency Bands (FDD)
E-UTRA Uplink (UL) operating band Downlink (DL) operating band
Operating BS receive UE transmit BS transmit UE receive Duplex Mode
Band FUL_low – FUL_high FDL_low – FDL_high
1 1920 MHz – 1980 MHz 2110 MHz – 2170 MHz FDD
5 824 MHz – 849 MHz 869 MHz – 894MHz FDD
9 1749.9 MHz – 1784.9 MHz 1844.9 MHz – 1879.9 MHz FDD
11 1427.9 MHz – 1452.9 MHz 1475.9 MHz – 1500.9 MHz FDD
20 832 MHz - 862 MHz 791 MHz - 821 MHz FDD
21 1447.9 MHz - 1462.9 MHz 1495.9 MHz - 1510.9 MHz FDD
22 3410 MHz - 3500 MHz 3510 MHz - 3600 MHz FDD


LTE/LTE-A Frequency Bands (TDD)
Uplink (UL) operating band Downlink (DL) operating band
E-UTRA
BS receive UE transmit BS transmit UE receive
Operating Duplex Mode
Band
FUL_low – FUL_high FDL_low – FDL_high

33 1900 MHz – 1920 MHz 1900 MHz – 1920 MHz TDD








3400 MHz – 3400 MHz –
41 TDD
3600MHz 3600MHz


Orthogonal Frequency Division Multiple Access

l OFDM is the modulation scheme for LTE in downlink and
uplink (as reference)
l Some technical explanation about our physical base: radio
link aspects


What does it mean to use the radio channel?
Using the radio channel means to deal with aspects like:
C
A

D

B
Transmitter Receiver

MPP

Time variant channel
Doppler effect

Frequency selectivity attenuation

Still the same “mobile” radio problem:
Time variant multipath propagation
A: free space
A: free space
B: reflection
B: reflection
C C: diffraction
C: diffraction
A
D: scattering
D: scattering
D

B
Transmitter Receiver

Multipath Propagation reflection: object is large
and Doppler shift compared to wavelength
scattering: object is
small or its surface
irregular


Multipath channel impulse response
The CIR consists of L resolvable propagation paths
L 1
h  , t    ai  t  e     i 
ji  t 

i 0

path attenuation path phase path delay

|h|²


delay spread


Radio Channel – different aspects to discuss

Bandwidth or

Wideband Narrowband

Symbol duration
or t
t
Short symbol Long symbol
duration duration
Channel estimation: Frequency?
Pilot mapping Time?

frequency distance of pilots? Repetition rate of pilots?

Frequency selectivity - Coherence Bandwidth
Here: substitute with single
power Scalar factor = 1-tap
Frequency selectivity

How to combat
channel influence?

f
Narrowband = equalizer
Can be 1 - tap

Wideband = equalizer
Here: find Must be frequency selective
Math. Equation
for this curve


Time-Invariant Channel: Scenario
Fixed Scatterer

ISI: Inter Symbol
Interference:
Happens, when
Delay spread >
Symbol time

Successive
Fixed Receiver
symbols Transmitter

will interfere Channel Impulse Response, CIR
Transmitter Receiver collision
Signal Signal

t t
Delay Delay spread

→time dispersive


Motivation: Single Carrier versus Multi Carrier
TSC
|H(f)| f
Source: Kammeyer; Nachrichtenübertragung; 3. Auflage

1
B
TSC
B

t
|h(t)|

t

l Time Domain
l Delay spread > Symboltime TSC
→ Inter-Symbol-Interference (ISI) → equalization effort

l Frequency Domain
l Coherence Bandwidth Bc < Systembandwidth B
→ Frequency Selective Fading → equalization effort


Motivation: Single Carrier versus Multi Carrier
TSC
|H(f)| f Source: Kammeyer; Nachrichtenübertragung; 3. Auflage

1
B
B TSC

t
|h(t)|

f t
|H(f)|
B 1
f  
B N TMC

t
TMC  N  TSC


What is OFDM?

Single carrier
transmission,
e.g. WCDMA

Broadband, e.g. 5MHz for WCDMA

Orthogonal
Frequency
Division
Multiplex
Several 100 subcarriers, with x kHz spacing


Idea: Wide/Narrow band conversion

ƒ

…
S/P

…
…

H(ƒ) t / Tb t / Ts
h(τ) h(τ)
„Channel
Memory“

τ τ
One high rate signal: N low rate signals:
Frequency selective fading Frequency flat fading


COFDM

Mapper
X
+
X

Data
with
OFDM
FEC Σ

.....
symbol
overhead
Mapper

X
+
X


OFDM signal generation
00 11 10 10 01 01 11 01 …. e.g. QPSK

h*(sinjwt + cosjwt) h*(sinjwt + cosjwt) => Σ h * (sin.. + cos…)

Frequency

time

OFDM
symbol
duration Δt 2012 |
November LTE Introduction | 25

Fourier Transform, Discrete FT
Fourier Transform

H ( f )   h(t )e 2 j ft
dt ;


h(t )   H ( f )e  2 j ft
df ;


Discrete Fourier Transform (DFT)

N 1 N 1 N 1
n n
H n   hk e 2 j k n / N
  hk cos(2  k )  j  hk sin(2  k );
k 0 k 0 N k 0 N
N 1  2 j k
n
1
hk 
N
H e
n 0
n
N
;


OFDM Implementation with FFT
(Fast Fourier Transformation)
Transmitter Channel
d(0)
Map

IDFT NFFT
d(1)

P/S
S/P

b (k ) Map
. s(n)
.
.
d(FFT-1)
Map

h(n)
Receiver
d(0)
Demap
n(n)
DFT NFFT

d(1)
P/S

ˆ k
S/P

b( ) Demap
.
r(n)
.
.
d(FFT-1)
Demap


Inter-Carrier-Interference (ICI)
10

SMC  f 
0

-10

-20


-30

xx
S
-40

-50

-60

-70
-1 -0.5 0 0.5 1
 

f-2 f-1 f0 f1 f2 f

Problem of MC - FDM ICI
Overlapp of neighbouring subcarriers
→ Inter Carrier Interference (ICI).
Solution
“Special” transmit gs(t) and receive filter gr(t) and frequencies fk allows orthogonal
subcarrier
→ Orthogonal Frequency Division Multiplex (OFDM)


Rectangular Pulse

A(f)

Convolution
sin(x)/x

t
f
Δt
Δf
time frequency


Orthogonality
Orthogonality condition: Δf = 1/Δt

Δf


ISI and ICI due to channel

Symbol l-1 l l+1

 h  n

n Receiver DFT
Window
Delay spread

fade in (ISI) fade out (ISI)


ISI and ICI: Guard Intervall

Symbol l-1 l l+1

 h  n TG  Delay Spread

n Receiver DFT
Window
Delay spread

Guard Intervall guarantees the supression of ISI!

Guard Intervall as Cyclic Prefix
Cyclic Prefix

Symbol l-1 l l+1

 h  n TG  Delay Spread

n Receiver DFT
Window
Delay spread

Cyclic Prefix guarantees the supression of ISI and ICI!

Synchronisation

Cyclic Prefix
OFDM Symbol : l  1 l l 1
CP CP CP CP

Metric

-
Search window

~
n

DL CP-OFDM signal generation chain
l OFDM signal generation is based on Inverse Fast Fourier Transform
(IFFT) operation on transmitter side:

Data QAM N Useful
1:N OFDM Cyclic prefix
source Modulator symbol IFFT N:1 OFDM
symbols insertion
streams symbols

Frequency Domain Time Domain

l On receiver side, an FFT operation will be used.


OFDM: Pros and Cons
Pros:
 scalable data rate
 efficient use of the available bandwidth
 robust against fading
 1-tap equalization in frequency domain

Cons:
 high crest factor or PAPR. Peak to average power ratio
 very sensitive to phase noise, frequency- and clock-offset
 guard intervals necessary (ISI, ICI) → reduced data rate


MIMO =
Multiple Input Multiple Output Antennas


MIMO is defined by the number of Rx / Tx Antennas
and not by the Mode which is supported Mode

1 1 SISO Typical todays wireless Communication System

Single Input Single Output

Transmit Diversity
1 1 MISO l Maximum Ratio Combining (MRC)
l Matrix A also known as STC
M
Multiple Input Single Output
l Space Time / Frequency Coding (STC / SFC)

Receive Diversity

1 1
SIMO l Maximum Ratio Combining (MRC)

Single Input Multiple Output Receive / Transmit Diversity
M
Spatial Multiplexing (SM) also known as:
l Space Division Multiplex (SDM)
l True MIMO
1 1 MIMO l Single User MIMO (SU-MIMO)
Multiple Input Multiple Output l Matrix B
M M
Space Division Multiple Access (SDMA) also known as:
l Multi User MIMO (MU MIMO)
l Virtual MIMO
Definition is seen from Channel l Collaborative MIMO
Multiple In = Multiple Transmit Antennas Beamforming


MIMO modes in LTE

-Spatial Multiplexing
-Tx diversity
-Multi-User MIMO
-Beamforming
-Rx diversity
Increased
Increased
Throughput per
Throughput at
Better S/N UE
Node B


Diversity – some thoughts
The SISO channel:
Fading on the air interface
h11
transmit signal s received signal r

j
r  h11  s  n  h11 e s  n

Amplitude phase
scaling rotation

The transmit signal is modified in amplitude and phase
plus additional noise


Diversity – some thoughts: matched filter
The SISO channel and matched filter (maximum ratio combining):

h11 h*11
received signal r estimated signal ř
transmit signal s

~  h* r  h* h s  h* n  h e  j h e j s  h* n  h 2 s  h* n
r 11 11 11 11 11 11 11 11 11

Idea: matched filter multiplies received signal with
conjugate of channel -> maximizes SNR

Transmitted signal s is estimated as:
~
r
~
s 2
h11

Diversity – some thoughts: performance of SISO
Euclidic distance
Detector
0 threshold
1
Decay with SNR, only
one channel available -
> fading will deteriorate
noise amplitude Bit error rate
distribution

Modulation Bit error Data rate = bits
scheme probability per symbol

BPSK 1/(4*SNR) 1

QPSK 1/(2*SNR) 2

16 QAM 5/(2*SNR) 4


RX Diversity

Maximum Ratio Combining depends on different fading of the
two received signals. In other words decorrelated fading
channels


Receive diversity gain – SIMO: 1*NRx
Receive antenna 1
time
r1=h11s+n1
Transmit s NTx NRx
h11
antenna Receive antenna 2
h12
r2=h12s+n2
h1NRx
Receive antenna NRx
rnrx=h1nrxs+nnrx
~  h* s  h* s  ...  h* s
r 11 1 12 2 1nRX nRX
Said:
  h11  h12  ...  h1nRx  s  h n  h n  ...  h
2 2 2
  * * *
nnRx diversity
  11 1 12 2 1nRx
order nRx
signal after maximum ratio combining
1
Pe ~
Probability for errors SNR nRx


TX Diversity: Space Time Coding
Fading on the air interface

data
The same signal is transmitted at differnet
antennas
space Aim: increase of S/N ratio 
increase of throughput
 s1  s2 
*
S2   * 
Alamouti Coding = diversity gain
time

approaches
 s2 s1  RX diversity gain with MRRC!
Alamouti Coding -> benefit for mobile communications


Space Time Block Coding according to Alamouti
(when no channel information at all available at transmitter)

d e1  h1  r1  h2  r2   h1 ²  h2 ²  d1  n'1
* *
From slide before:

d e 2  h2  r1  h1  r2   h1 ²  h2 ²  d 2  n'2
* *

Probability for errors (Alamouti) 1
Pe ~
SNR 2
Compare
with Rx diversity

Alamouti coding is full diversity gain
and full rate, but it only works for 1
Pe ~
2 antennas SNR nRx
(due to Alamouti matrix is orthogonal)


MIMO Spatial Multiplexing
C=B*T*ld(1+S/N)
SISO:
Single Input
Single Output

Higher capacity without additional spectrum!
MIMO: S
C   T  B  ld (1  ) ?
min( N T , N R )
i
i
i 1
N
Multiple Input i

Multiple Output

Increasing
capacity per cell


Spatial multiplexing – capacity aspects
Ergodic mean capacity of a SISO channel calculated as:

h11
2
C  EH {log 2 (1   h11 )}
r  s  h11  n
Received signal r with
sent signal s, channel h11 and
AWGN with σ=n
P
 
represents the signal to noise ratio SNR
2
at the receiver branch

Or simplified:  S With B = bandwidth
C  B * log 2 1   and S/N = signal to

 N
noise ratio


Ergodic mean capacity of a MIMO channel is even worse 


    
H  
C  EH l og 2 det  I nR  HH  
 

   nT  
 n1 n1

n2 n2

InR is an Identity matrix with size nT x nR nT nR

r  sH n
HH is the Hermetian complex
Received signal r with
sent signal s, channel H and
AWGN with σ=n


Some theoretical ideas:


    
H  
C  EH l og 2 det  I nR  HH    EH nR * log 2 1   
 

   nT  


We increase to number of HH H
lim  I nR
transmit antennas to ∞, and see: nT  nT

So the result is, if the number of Tx antennas is infinity, the
capacity depends on the number of Rx antennas:

After this heavy mathematics the result: If we increase the
number of Tx and Rx antennas, we can increase the capacity!

The MIMO promise
l Channel capacity grows linearly with antennas 

Max Capacity ~ min(NTX, NRX)

l Assumptions 
l Perfect channel knowledge
l Spatially uncorrelated fading

l Reality 
l Imperfect channel knowledge
l Correlation ≠ 0 and rather unknown


Spatial Multiplexing
Coding Fading on the air interface

data

data

Throughput: <200%
200%
100%

Spatial Multiplexing: We increase the throughput
but we also increase the interference!

MIMO – capacity calculations, e.g. 2x2 MIMO
n1
h11
s1 r1
h12
n2
h21
s2 r2
h22
This results in the equations:
Or as matrix:
r1 = s1*h11 + s2*h21 + n1
 r1   h11 h12   s1   n1 
r2 = s2*h22 + s1*h12 + n2 r   h  *  s   n 
 2   21 h22   2   2 
100%
General written as: r = s*H +n
To solve this equation, we have to know H

Introduction – Channel Model II Correlation of
propagation
h11
pathes
h21
s1 r1
hMR1
h12
s2 h22 r2 estimates
Transmitter hMR2 Receiver
h1MT h2MT
NTx NRx
sNTx hMRMT rNRx
antennas antennas

s H r

Rank indicator

Capacity ~ min(NTX, NRX) → max. possible rank!
But effective rank depends on channel, i.e. the
correlation situation of H


Spatial Multiplexing prerequisites
Decorrelation is achieved by:

l Decorrelated data content on each spatial stream difficult

l Large antenna spacing Channel
condition

l Environment with a lot of scatters near the antenna
(e.g. MS or indoor operation, but not BS)
Technical
l Precoding assist
But, also possible
that decorrelation
l Cyclic Delay Diversity is not given


MIMO: channel interference + precoding

MIMO channel models: different ways to combat against
channel impact:

I.: Receiver cancels impact of channel

II.: Precoding by using codebook. Transmitter assists receiver in
cancellation of channel impact

III.: Precoding at transmitter side to cancel channel impact


MIMO: Principle of linear equalizing
R = S*H + n

Transmitter will send reference signals or pilot sequence
to enable receiver to estimate H.

n H-1
Rx
s r ^
r
Tx H
LE

The receiver multiplies the signal r with the
Hermetian conjugate complex of the transmitting
function to eliminate the channel influence.

Linear equalization – compute power increase

h11 H = h11

SISO: Equalizer has to estimate 1 channel

h11
h12 h11 h12
H=
h21 h22
h21 h22

2x2 MIMO: Equalizer has to estimate 4 channels


transmission – reception model
noise

s + r
A H R

transmitter channel receiver

•Modulation, •detection,
•Power •estimation
•„precoding“, •Eliminating channel
•etc. Linear equalization impact
at receiver is not •etc.
very efficient, i.e.
noise can not be cancelled


MIMO – work shift to transmitter

Channel Receiver
Transmitter


MIMO Precoding in LTE (DL)
Spatial multiplexing – Code book for precoding

Code book for 2 Tx:
Codebook Number of layers 
index
1 2 Additional multiplication of the
1  1 1 0
0 0 
 
 
2 0 1 
layer symbols with codebook
0  1 1 1  entry
1 1   
  2 1 1
1 1 1 1 1 
2   
2 1 2  j  j
1 1
3   -
2 1
1 1 
4   -
2  j
1 1
5   -
2  j 


MIMO precoding
precoding
Ant1
Ant2 t
+
 1
2 1 ∑

t
+
1
precoding -1
1

∑=0
t
t


MIMO – codebook based precoding
Precoding
codebook

noise

s + r
A H R

transmitter channel receiver

Precoding Matrix Identifier, PMI

Codebook based precoding creates
some kind of „beamforming light“

MIMO: avoid inter-channel interference – future outlook

e.g. linear precoding:
V1,k Y=H*F*S+V

S Link adaptation
+
Transmitter H Space time
F receiver

xk +
yk
VM,k

Feedback about H

Idea: F adapts transmitted signal to current channel conditions

MAS: „Dirty Paper“ Coding – future outlook

l Multiple Antenna Signal Processing: „Known Interference“
l Is like NO interference
l Analogy to writing on „dirty paper“ by changing ink color accordingly

„Known
„Known „Known „Known
Interference
Interference Interference Interference
is No
is No is No is No
Interference“
Interference“ Interference“ Interference“


Spatial Multiplexing
Codeword Fading on the air interface

data

Codeword
data

Spatial Multiplexing: We like to distinguish the 2 useful
Propagation passes:
How to do that? => one idea is SVD

Idea of Singular Value Decomposition
s1 MIMO r1
know

r=Hs+n s2 r2

channel H
Singular Value
Decomposition
~
s1 ~
r1
SISO

wanted ~ ~
s2 r2
~ ~ ~
r=Ds+n
channel D

Singular Value Decomposition (SVD)

h11 h12
r= H s + n
h21 h22
h11 h0
d1 12
r= U H (V*)T s + n
h0 h22
21 d2

d1 0
(U*)T r= (U*)T U D (V*)T s + (U*)T n
0 d2

~ = (U*)T U d1 0 ~ ~
(U*)T r D (V*)T s + (U*)T n
0 d2


Singular Value Decomposition (SVD)
r=Hs+n

H = U Σ (V*)T
U = [u1,...,un] eigenvectors of (H*)T H
V = [v1,...,vm] eigenvectors of H (H*)T

 1 0 0 
0  i eigenvalues of (H*)T H
 0 0

 2

 0 0 3  singular values  i  i
 0
 0  ~ = (U*)T r
r
~
~= Σ s + n
r ~ ~ s = (V*)T s
~ = (U*)T n
n

MIMO and singular value decomposition SVD
Real channel n1
h11
s1 r1
h12
n2
h21
s2 r2
h22
Channel model with SVD
n1
s1 σ1 r1
U Σ VH n2
s2 σ2 r2

SVD transforms channel into k parallel AWGN channels


MIMO: Signal processing considerations
MIMO transmission can be expressed as
r = Hs+n which is, using SVD = UΣVHs+n
n1
s1 σ1 r1
V U Σ VH n2 UH
s2 σ2 r2
Imagine we do the following:
1.) Precoding at the transmitter:
Instead of transmitting s, the transmitter sends s = V*s
2.) Signal processing at the receiver
Multiply the received signal with UH, r = r*UH

So after signal processing the whole signal can be expressed as:
r =UH*(UΣVHVs+n)=UHU Σ VHVs+UHn = Σs+UHn
=InTnT =InTnT

MIMO: limited channel feedback
Transmitter H Receiver
n1
s1 σ1 r1
V U Σ VH n2 UH
s2 σ2 r2

Idea 1: Rx sends feedback about full H to Tx.
-> but too complex,
-> big overhead
-> sensitive to noise and quantization effects

Idea 2: Tx does not need to know full H, only unitary matrix V
-> define a set of unitary matrices (codebook) and find one matrix in the codebook that
maximizes the capacity for the current channel H
-> these unitary matrices from the codebook approximate the singular vector structure
of the channel
=> Limited feedback is almost as good as ideal channel knowledge feedback


Cyclic Delay Diversity, CDD
A2
A1 Amp
litud
D
e

Transmitter B

Delay Spread Time
Delay

Multipath propagation
precoding

+

 +
precoding Time
No multipath propagation Delay


„Open loop“ und „closed loop“ MIMO
Open loop (No channel knowledge at transmitter)

r  Hs  n Channel
Status, CSI

Rank indicator

Closed loop (With channel knowledge at transmitter

r  HWs  n Channel
Status, CSI

Rank indicator
Adaptive Precoding matrix („Pre-equalisation“)
Feedback from receiver needed (closed loop)


MIMO transmission modes
Transmission mode2
Transmission mode3
TX diversity
Transmission mode1 Open-loop spatial
SISO multiplexing

7 transmission
Transmission mode4 Transmission mode7
Closed-loop spatial modes are SISO, port 5
multiplexing defined = beamforming in TDD

Transmission mode6
Transmission mode5 Closed-loop
Multi-User MIMO spatial multiplexing,
using 1 layer

Transmission mode is given by higher layer IE: AntennaInfo

the classic:
1Tx + 1RX
Transmission mode1
antenna
SISO

PDCCH indication via
DCI format 1 or 1A

PDSCH transmission via
single antenna port 0 No feedback regarding
antenna selection or
precoding needed


Transmission mode 2
Transmit
diversity PDCCH indication via
DCI format 1 or 1A
Codeword is sent
redundantly over several
streams
1 codeword

PDSCH transmission via
2 Or 4 antenna ports No feedback regarding
precoding needed


MIMO transmission modes No feedback regarding
Transmission mode 3 precoding needed
Transmit diversity or Open loop
spatial multiplexing
DCI format 1A

1 codeword PDSCH transmission
Via 2 or 4 antenna ports

DCI format 2A

1 2
codeword codewords

PMI feedback
PDSCH spatial multiplexing PDSCH spatial multiplexing, using CDD
with 1 layer

MIMO transmission modes Closed loop MIMO =
UE feedback needed regarding
Transmission mode 4 precoding and antenna
Transmit diversity or Closed loop selection
spatial multiplexing
DCI format 1A


DCI format 2
precoding

precoding
1 2
codeword codewords

PMI feedback PMI feedback
PDSCH spatial multiplexing PDSCH spatial multiplexing
with 1 layer

Transmission mode 5
Transmit diversity or
Multi User MIMO
DCI format 1A


PUSCH
UE1
Codeword
PDCCH indication
via DCI format 1D

UE2 PDSCH multiplexing to several UEs.
Codeword PUSCH multiplexing in Uplink


MIMO transmission modes Closed loop MIMO =
UE feedback needed regarding
Transmission mode 6 precoding and antenna
Transmit diversity or Closed loop selection
spatial multiplexing with 1 layer
DCI format 1A

via 2 or 4 antenna ports

DCI format 1B

Codeword is split into
1
streams, both streams have
codeword
to be combined

feedback
PDSCH spatial multiplexing, only 1 codeword


Transmission mode 7
Transmit diversity or beamforming

DCI format 1A

via 1, 2 or 4 antenna ports

DCI format 1

1
codeword

PDSCH sent over antenna port 5 = beamforming


Beamforming
Closed loop precoded
Adaptive Beamforming beamforming

•Classic way •Kind of MISO with channel
knowledge at transmitter
•Antenna weights to adjust beam
•Precoding based on feedback
•Directional characteristics
•No specific antenna
•Specific antenna array geometrie
array geometrie
•Dedicated pilots required •Common pilots are sufficient


Spatial multiplexing vs beamforming

Spatial multiplexing increases throughput, but looses coverage


Spatial multiplexing vs beamforming

Beamforming increases coverage


Basic OFDM parameter
LTE
1
f  15 kHz 
T
Fs  N FFT  f
N FFT
Fs   3.84Mcps
256
f
NFFT  2048

Coded symbol rate= R

Sub-carrier CP
S/P Mapping IFFT insertion

N TX Data symbols

Size-NFFT


LTE Downlink:
Downlink slot and (sub)frame structure
Symbol time, or number of symbols per time slot is not fixed

One radio frame, Tf = 307200Ts=10 ms

One slot, Tslot = 15360Ts = 0.5 ms

#0 #1 #2 #3 #18 #19

One subframe

We talk about 1 slot, but the minimum resource is 1 subframe = 2 slots !!!!!

Ts  1 15000  2048
Ts = 32.522 ns


Resource block definition
1 slot = 0,5msec

Resource block
=6 or 7 symbols
In 12 subcarriers
12 subcarriers

Resource element

DL UL
N symb or N symb
6 or 7,
Depending on
cyclic prefix


LTE Downlink
OFDMA time-frequency multiplexing
frequency
QPSK, 16QAM or 64QAM modulation

UE4
1 resource block =
180 kHz = 12 subcarriers UE5

UE3
UE2

UE6
Subcarrier spacing = 15 kHz time
UE1

1 subframe =
*TTI = transmission time interval 1 slot = 0.5 ms = 1 ms= 1 TTI*=
** For normal cyclic prefix duration 7 OFDM symbols** 1 resource block pair


LTE: new physical channels for data and control
Physical Control Format Indicator Channel PCFICH:
Indicates Format of PDCCH

Physical Downlink Control Channel PDCCH:
Downlink and uplink scheduling decisions

Physical Downlink Shared Channel PDSCH: Downlink data

Physical Hybrid ARQ Indicator Channel PHICH:
ACK/NACK for uplink packets

Physical Uplink Shared Channel PUSCH: Uplink data

Physical Uplink Control Channel PUCCH:
ACK/NACK for downlink packets, scheduling requests, channel quality info


LTE Downlink: FDD channel mapping example

Subcarrier #0 RB

LTE – spectrum flexibility

l LTE physical layer supports any bandwidth from 1.4 MHz
to 20 MHz in steps of 180 kHz (resource block)
l Current LTE specification supports only a subset of 6
different system bandwidths
l All UEs must support the maximum bandwidth of 20 MHz

Channel Bandwidth [MHz]

Channel Transmission Bandwidth Configuration [RB]

bandwidth Transmission
BWChannel 1.4 3 5 10 15 20 Bandwidth [RB]

Channel edge
Channel edge

[MHz] Resource block

FDD and
TDD mode 6 15 25 50 75 100

number of resource blocks
Active Resource Blocks DC carrier (downlink only)


LTE Downlink:
baseband signal generation

code words layers antenna ports

Modulation OFDM OFDM signal
Scrambling
Mapper Mapper generation
Layer
Precoding
Mapper
Modulation OFDM OFDM signal
Scrambling
Mapper Mapper generation

1 stream =
several
Avoid QPSK For MIMO Weighting 1 OFDM subcarriers,
constant 16 QAM Split into data symbol per based on
sequences 64 QAM Several streams for stream Physical
streams if MIMO ressource
needed blocks

Adaptive modulation and coding

Transportation block size

User data FEC

Flexible ratio between data and FEC = adaptive coding


Channel Coding Performance


Automatic repeat request, latency aspects
•Transport block size = amount of
data bits (excluding redundancy!)
•TTI, Transmit Time Interval = time
duration for transmitting 1 transport
block

Transport block
Round
Trip
Time
ACK/NACK

Network UE

Immediate acknowledged or non-acknowledged
feedback of data transmission

HARQ principle: Stop and Wait

Δt = Round trip time

Tx Data Data Data Data Data Data Data Data Data Data
ACK/NACK
Demodulate, decode, descramble,
Rx FFT operation, check CRC, etc.
process
Processing time for receiver

Described as 1 HARQ process


HARQ principle: Multitasking

Δt = Round trip time

Tx Data Data Data Data Data Data Data Data Data Data
ACK/NACK
Demodulate, decode, descramble,
Rx FFT operation, check CRC, etc.
process
ACK/NACK
Processing time for receiver

Rx Demodulate, decode, descramble,
process FFT operation, check CRC, etc.

t
Described as 1 HARQ process


LTE Round Trip Time RTT
n+4 n+4 n+4

ACK/NACK
PDCCH

PHICH
Downlink

HARQ
Data

Data
UL

Uplink

t=0 t=1 t=2 t=3 t=4 t=5 t=6 t=7 t=8 t=9 t=0 t=1 t=2 t=3 t=4 t=5

1 frame = 10 subframes

8 HARQ processes
RTT = 8 msec


HARQ principle: Soft combining

lT i is a e am l o h n e co i g

Reception of first transportation block.
Unfortunately containing transmission errors



l hi i n x m le f cha n l c ing

Reception of retransmitted
transportation block.
Still containing transmission errors


1st transmission with puncturing scheme P1
l T i is a e am l o h n e co i g

2nd transmission with puncturing scheme P2
l hi i n x m le f cha n l c ing

Soft Combining = Σ of transmission 1 and 2
l Thi is an exam le of channel co ing

Final decoding
lThis is an example of channel coding


Hybrid ARQ
Chase Combining = identical retransmission
Turbo Encoder output (36 bits)

Systematic Bits
Parity 1
Parity 2
Transmitted Bit Rate Matching to 16 bits (Puncturing)
Original Transmission Retransmission
Systematic Bits
Parity 1
Parity 2

Punctured Bit Chase Combining at receiver
Systematic Bits
Parity 1
Parity 2


Hybrid ARQ
Incremental Redundancy
Turbo Encoder output (36 bits)

Systematic Bits
Parity 1
Parity 2

Rate Matching to 16 bits (Puncturing)
Original Transmission Retransmission
Systematic Bits
Parity 1
Parity 2

Punctured Bit Incremental Redundancy Combining at receiver
Systematic Bits
Parity 1
Parity 2


LTE Physical Layer:
SC-FDMA in uplink

Single Carrier Frequency Division
Multiple Access


LTE Uplink:
How to generate an SC-FDMA signal in theory?
Coded symbol rate= R

Sub-carrier CP
DFT Mapping IFFT insertion

NTX symbols

Size-NTX Size-NFFT

 LTE provides QPSK,16QAM, and 64QAM as uplink modulation schemes
 DFT is first applied to block of NTX modulated data symbols to transform them into
frequency domain
 Sub-carrier mapping allows flexible allocation of signal to available sub-carriers
 IFFT and cyclic prefix (CP) insertion as in OFDM
 Each subcarrier carries a portion of superposed DFT spread data symbols
 Can also be seen as “pre-coded OFDM” or “DFT-spread OFDM”


LTE Uplink:
How does the SC-FDMA signal look like?

 In principle similar to OFDMA, BUT:
 In OFDMA, each sub-carrier only carries information related to one specific symbol
 In SC-FDMA, each sub-carrier contains information of ALL transmitted symbols


LTE uplink
SC-FDMA time-frequency multiplexing
1 resource block =
180 kHz = 12 subcarriers Subcarrier spacing = 15 kHz

frequency

UE1 UE2 UE3
1 slot = 0.5 ms =
7 SC-FDMA symbols**
1 subframe =
1 ms= 1 TTI*
UE4 UE5 UE6

*TTI = transmission time interval
** For normal cyclic prefix duration

time QPSK, 16QAM or 64QAM modulation


LTE Uplink:
baseband signal generation
UE specific
Scrambling code

Modulation Transform Resource SC-FDMA
Scrambling element mapper
mapper precoder signal gen.

Mapping on
physical 1 stream =
Discrete
Ressource, several
Avoid QPSK Fourier
i.e. subcarriers,
constant 16 QAM Transform
subcarriers based on
sequences 64 QAM not used for Physical
(optional) reference ressource
signals blocks

LTE Protocol Architecture


LTE Protocol Architecture
Reduced complexity

l Reduced number of transport channels

l Shared channels instead of dedicated channels

l Reduction of Medium Access Control (MAC) entities

l Streamlined concepts for broadcast / multicast (MBMS)

l No inter eNodeB soft handover in downlink/uplink

l No compressed mode

l Reduction of RRC states


EUTRAN stack: protocol layers overview
EMM ESM User plane

Radio Resource Control
RRC

Packet Data Convergence
PDCP
Control & Measurements

Radio Bearer

Radio Link Control
RLC
Logical channels
Medium Access Control
MAC
Transport channels

PHYSICAL LAYER


User plane Header compression (ROHC)
In-sequence delivery at handover
Duplicate detection
Ciphering for user/control plane
Integrity protection for control plane
Timer based SDU discard in Uplink…

UE eNB
AM, UM, TM
ARQ
PDCP PDCP
(Re-)segmentation
Concatenation
RLC RLC In-sequence delivery
Duplicate detection
MAC MAC SDU discard
Reset…
PHY PHY

Mapping between logical and PDCP = Packet Data Convergence Protocol
transport channels RLC = Radio Link Control
(De)-Multiplexing MAC = Medium Access Control
Traffic volume measurements PHY = Physical Layer
HARQ SDU = Service Data Unit
Priority handling (H)ARQ = (Hybrid) Automatic Repeat Request
Transport format selection…


Control plane Broadcast
Paging
RRC connection setup
Radio Bearer Control
Mobility functions
UE measurement control…

UE eNB MME

NAS NAS

RRC RRC

PDCP PDCP

RLC RLC
EPS bearer management
Authentication
MAC MAC
ECM_IDLE mobility handling
Paging origination in ECM_IDLE
Security control…
PHY PHY

EPS = Evolved packet system
RRC = Radio Resource Control
NAS = Non Access Stratum
ECM = EPS Connection Management

EPS Bearer Service Architecture

E-UTRAN EPC Internet

UE eNB S-GW P-GW Peer
Entity

End-to-end Service

EPS Bearer External Bearer

Radio Bearer S1 Bearer S5/S8 Bearer

Radio S1 S5/S8 Gi


Channel structure: User + Control plane
Protocol structure


LTE channel mapping


LTE – channels
MTCH MCCH CCCH DCCH DTCH PCCH BCCH
DL logical channels

MCH DL-SCH PCH BCH
DL transport channels

DL physical channels
PMCH PCFICH PDCCH PDSCH PHICH PBCH

CCCH DCCH DTCH
UL logical channels

RACH UL-SCH
UL transport channels

UL physical channels
PRACH PUCCH PUSCH


LTE – uplink channels
Mapping between logical and transport channels

CCCH DCCH DTCH
Uplink
Logical channels

Uplink
Transport channels
RACH UL-SCH


LTE resource allocation principles


LTE resource allocation
Scheduling of downlink and uplink data
Check PDCCH for your UE ID. You may
find here Uplink and/or Downlink Physical Uplink Shared Channel
resource allocation information (PUSCH)

Physical Downlink Control Physical Downlink Shared
Channel (PDCCH) Channel (PDSCH)

Physical Control Format I would like to receive data on PDSCH
Indicator Channel (PCFICH), and / or send data on PUSCH
Info about PDCCH format

?


Resource allocation types in LTE
Allocation type DCI Format Scheduling Antenna
Type configuration
Type 0 / 1 DCI 1 PDSCH, one SISO,
codeword TxDiversity
DCI 2A PDSCH, two MIMO, open
codewords loop
DCI 2 PDSCH, two MIMO, closed
codewords loop
Type 2 DCI 0 PUSCH SISO

DCI 1A PDSCH, one SISO,
codeword TxDiversity
DCI 1C PDSCH, very SISO
compact
codeword


Type 0 and 1 for distributed allocation in frequency domain
Channel bandwidth

f
Type 2 for contiguous allocation in frequency domain
Channel bandwidth
Transmission bandwidth

f

Resource Block Group
Reminder: 1 resource block =
12 subcarriers in frequency domain
f

Resource allocation is performed
based on resource block groups.
1 resource block group may consist of
1, 2, 3 or 4 resource blocks

Resource block groups,
RBG sizes


Resource allocation type 0
Type 0 (for distributed frequency allocation of Downlink resource,
SISO and MIMO possible)

l Bitmap to indicate which resource block groups, RBG are allocated
l One RBG consists of 1-4 resource blocks: Channel RBG size P
bandwidth
≤10 1
l Granularity is RBG size 11-26 2
27-63 3
l Number of resource block groups NRBG
64-110 4
is given as:

N RBG  N RB / P
DL

l Allocation bitmap has same length than NRBG


Resource allocation type 0 example
Calculation example for type 0:
l Channel bandwidth = 10MHz
l -> 50 resource blocks
l -> Resource block group RBG size = 3
l -> bitmap size = 17

if N RB mod P  0
DL
then one of the RBGs is of size
DL

N RB  P  N RB / P
DL

i.e. here 50 mod 3 = 16, so the last resource block group has the size 2.

-> some allocations are not possible, e.g. here you can allocate
48 or 50 resource blocks, but not 49!


N RBG  N RB / P
DL
 = round up, i.e.3.5 = 4 reminder

 DL
N RB / P  = round down, i.e. 3.49 = 3


Resource allocation type 0: example
Channel bandwidth = 10MHz -> 50 RBs -> RBG size = 3 -> number of RBGs = 17

RBG#0 RBG#1 RBG#6 RBG#NRBG-1

0 1 2 3 16 17 18 19 20 21 22 23 N RB  3 N RB  2 N RB  1
DL DL DL

Allocation bitmap (17bit): 10000011000000001

Granularity: 1 bit allocates 1 ressource block group ´f

Type 1 (for distributed frequency allocation of Downlink resource,
SISO and MIMO possible)

One Resource Block Group
consists of 1-4 resource blocks: Channel RBG size P
bandwidth
l RBs are divided into log 2 ( P)
 ≤10 1
RBG subsets
11-26 2
l Granularity is resource block 27-63 3
64-110 4

l Bitmap indicates RBs inside a RBG subset allocated to the UE

l Resource block assignment consists of 3 fields:
l Field to indicate the selected RBG
l Field to indicate a shift of the resource allocation
l Field to indicate the specific RB within a RBG subset



RBG#0 RBG#1 RBG#NRBG-1

0 1 2 3 16 17 18 19 20 21 22 23 N RB  3 N RB  2 N RB  1
DL DL DL

Size 18 resource blocks
RBG subset #0
RBG#0 RBG#3 RBG#6 RBG#9 RBG#12 RBG#15

RBG#1 RBG#4 RBG#7 RBG#10 RBG#13 RBG#16 RBG subset #1

RBG#2 RBG#5 RBG#8 RBG#11 RBG#14 RBG subset #2

P= Number of RBG subsets with length:  N RB  1
DL
 N DL  1
 2 P P , p   RB  mod P
 P   P 

log 2 ( P)
 DL
 N  1  N RB  1
DL
N RB subset ( p)   RB 2   P  ( N RB  1) mod P  1
RBG DL
,p  mod P
 P   P 
 DL
 N RB  1  P

 N DL  1
, p   RB  mod P
indicate subset  P 
2
Number of bits to   P 



RBG#0 RBG#1 RBG#NRBG-1

0 1 2 3 16 17 18 19 20 21 22 23 N RB  2 N RB  1
DL DL

RBG#0 RBG#3 RBG#6 RBG#9 RBG#12 RBG#15 RBG#2 RBG#5 RBG#8 RBG#11 RBG#14
RBG subset #0 RBG subset #2

3 4 5 12 13 14 21 22 23 30 31 32 39 40 41 48 49 Resource blocks
assignment
Field 1: RBG subset selection Field 2: offset shift indication Field 3: resource block allocation

Allocation bitmap (17bit): 01 0 00011000000001
Size 14 bits
RBG subset#1 is selected

Bit = 0, no shift


The meaning of the shift offset bit:
Number of resource blocks in one RBG subset is bigger than the allocation bitmap
-> you can not allocate all the available resource blocks
-> offset shift to indicate which RBs are assigned
17 resource blocks belonging to the RBG subset#1

4 5 12 13 14 21 22 23 30 31 32 39 40 41 48 49 Resource blocks
3
assignment

14 bits in allocation bitmap
Bit = 0, no shift

The meaning of the shift offset bit:
Number of resource blocks in one RBG subset is bigger than the allocation bitmap
-> you can not allocate all the available resource blocks
-> offset shift to indicate which RBs are assigned
17 resource blocks belonging to the RBG subset #1

3 4 5 1 14 21 22 23 30 31 32 39 40 41 48 49 Resource blocks
12 3 assignment

14 bits in allocation bitmap
Bit = 1, offset shift

Type 0 allows the allocation based on resource block groups granularity
Channel bandwidth
Example: RBG size = 3 RBs

f
Type 1 allows the allocation based on resource block granularity
Channel bandwidth
1 resource block, RB

f

Localized mode

Starting resource block: RBStart
LCRB, length of contiguously allocated RBs

Number of allocated Resource blocks NRB

Channel bandwidth
Transmission bandwidth

RB#0
f
Resource block offset


Resource indication value, RIV
Type 0: bitmap used to indicate the resource allocation:
N RB / P = Length of allocation bitmap (17bit): 10000011000000001
DL

RIV = bin to
Type 1: bitmap used to indicate the resource allocation, with 3 fields: dec
conversion
Resource block group subset, shift indictor + resource allocation
Type 2: TS 36.213 section 7.1.6.3. gives formula to calculate RIV:
if ( LCRBs  1)  N RB / 2
DL

then

RIV  N RB ( LCRBs  1)  RBstart
DL

else

RIV  N RB ( N RB  LCRBs  1)  ( N RB  1  RBstart )
DL DL DL


Resource indication value, RIV in type 2 allocation
How to calculate the RIV value in allocation type 2, according to TS 36.213
Assumptions and given:
Starting resource block: RBStart=5 Localized mode, RBStart = 5, LCRB = 20
NDLRB = 50
LCRB, length of contiguously allocated RBs=20

f

RIV  N DL
RB ( LCRBs  1)  RBstart Formula from TS
36.213

Here:
RIV = 50 * (20-1) + 5
RIV = 955


Benefit of localized or distributed mode
„static UE“: frequency selectivity is not time variant -> localized allocation

Multipath causes frequency
Selective channel,
It can be time variant or
Non-time variant

„high velocity UE“: frequency selectivity is time variant -> distributed allocation


Resource allocation Uplink
Allocation type DCI Format Scheduling Antenna
Type configuration
Type 0 / 1 DCI 1 PDSCH, one codeword SISO, TxDiversity

DCI 2A PDSCH, two MIMO, open loop
LTE Uplink uses codewords
Type 2 allocation DCI 2 PDSCH, two MIMO, closed loop
codewords

Type 2 DCI 0 PUSCH SISO
DCI 1A PDSCH, one codeword SISO, TxDiversity

DCI 1C PDSCH, very compact SISO
codeword

Starting resource block: RBStart LCRB, length of contiguously allocated RBs

RB#0
Channel bandwidth

LTE Uplink: allocation of UL ressource
Scheduled UL UL bandwidth
bandwidth configuration

2 3 5
PUSCH
M RB 2 3 5  UL
N RB
Scheduled number of ressource blocks in UL must fullfill
formula above(αx are integer). Possible values are:

1 2 3 4 5 6 8 9 10 12

15 16 18 20 24 25 27 30 32 36

40 45 48 50 54 60 64 72 75 80

81 90 96 100


The LTE evolution Rel-9
eICIC
enhancements
Relaying
In-device Rel-10
Diverse Data co-existence
Application CoMP
Rel-11
Relaying
eICIC
eMBMS
enhancements SON
enhancements

MIMO 8x8 MIMO 4x4
Carrier Enhanced
Aggregation SC-FDMA

Public Warning
Positioning System Home eNodeB

Self Organizing
eMBMS Networks

DL UL
Multi carrier /
Dual Layer DL UL Multi-RAT
Beamforming Base Stations
LTE Release 8
FDD / TDD


What are antenna ports?
l 3GPP TS 36.211(Downlink)
“An antenna port is defined such that the channel over which a symbol on the
antenna port is conveyed can be inferred from the channel over which another
symbol on the same antenna port is conveyed.”

l What does that mean?

l The UE shall demodulate a received signal – which is transmitted over a
certain antenna port – based on the channel estimation performed on the
reference signals belonging to this (same) antenna port.



l Consequences of the definition
l There is one sort of reference signal per antenna port
l Whenever a new sort of reference signal is introduced by 3GPP (e.g. PRS),
a new antenna port needs to be defined (e.g. Antenna Port 6)

l 3GPP defines the following antenna port / reference signal
combinations for downlink transmission:
l Port 0-3: Cell-specific Reference Signals (CS-RS)
l Port 4: MBSFN-RS
l Port 5: UE-specific Reference Signals (DM-RS): single layer (TX mode 7)
l Port 6: Positioning Reference Signals (PRS)
l Port 7-8: UE-specific Reference Signals (DM-RS): dual layer (TX mode 8)
l Port 7-14: UE specific Referene Signals for Rel. 10
l Port 15 – 22: CSI specific reference signals, channel status info in Rel. 10


l Mapping „Antenna Port“ to „Physical Antennas“
Antenna Port Physical Antennas

1
AP0 PA0
AP1 1 W5,0
AP2 1 PA1
W5,1
AP3 1
PA2
AP4 … W5,2

W5,3
AP5 PA3
…
AP6

AP7 …
AP8 …

The way the "logical" antenna ports are mapped to the "physical" TX antennas lies
completely in the responsibility of the base station. There's no need for the base station
to tell the UE.


LTE antenna port definition
Antenna ports are linked to the reference signals
-> one example:

Normal CP

Cell Specific RS
PA -RS
PCFICH /PHICH /PDCCH
PUSCH or No Transmission

UE in connected mode, scans
UE in idle mode, scans for Positioning RS on antenna port 6
Antenna port 0, cell specific RS to locate its position


eMBMS
Physical Layer Scenarios

l Dedicated and mixed mode.
l Dedicated: carrier is only for MBMS = Single-cell MBMS.
l MBMS/Unicast mixed mode: MBMS and user data are transmitted using
time division duplex. Certain subframes carry MBMS data.
l Dedicated mode (single-cell scenario) offers use of new
subscarrier spacing, longer cyclic prefix (CP), 3 OFDM symbols.
OFDM Sub- Cyclic Prefix Length Cyclic Prefix
Configuration
Symbols carrier in Samples Length in µs
Normal CP 160 for 1st symbol 5.2 for 1st symbol
7
∆f = 15 kHz 144 for other symbols 4.7 for other symbols
12
Extended CP
6 512 16.7
∆f = 15 kHz
Extended CP eMBMS
3 24 1024 33.3
∆f = 7.5 kHz (Single cell scenario)


Resource block definition for MBMS
Δf=1/TSYMBOL=15kHz
1 Resource
Block =
12 subcarriers

f
f0 f1 f2
1 2 3 4 5 6 7 7 OFDM symbols

OFDM OFDM OFDM OFDM OFDM OFDM 6 OFDM symbols
CP Symbol CP Symbol CP Symbol CP Symbol CP Symbol CP Symbol

1 Resource Δf=
Block = 7.5kHz
24 subcarriers
For
f MBMS
OFDM OFDM OFDM
CP Symbol CP Symbol CP Symbol only!

Multimedia Broadcast Messaging Services, MBMS

Broadcast: Unicast:
Public info for Private info for dedicated user
everybody

Multicast:
Common info for
User after authentication


LTE MBMS architecture


MBMS in LTE

MBMS
MME
GW
|

M3 MBMS GW: MBMS Gateway
MCE: Multi-Cell/Multicast Coordination Entity
M1
|

MCE M1: user plane interface
M2: E-UTRAN internal control plane interface
M2 M3: control plane interface between E-UTRAN and EPC
|

eNB
Logical architecture for MBMS


MBMS broadcast service provision

Service announcement

Session Start

MBMS notification

Data transfer

Session Stop

See 3GPP TS23.246, Section 4.4.3


MBMS broadcast service provision
time

UE1
Local service de -
UE local
service
activation
How is the activation

UE2 UE informed about all
Idle period Data Data
this in detail? Data
of seconds transfer transfer transfer

Broadcast session start Session stop
Start Broadcast Stop
Service announcment
announcement t

Service 1 session1
Service 1 Session 2

Broadcast service Announcement


MBMS Signaling

l New logical Channels: POC VoIP

FTP WAP HTTP MMS SIP

l MBMS point-to-multipoint Control TCP / UDP
IP
Channel (MCCH)
l MBMS point-to-multipoint Traffic SNSM

Channel (MTCH) SM SMREG TC CTC

l MBMS point-to-multipoint Scheduling GMMSM MMTC

GMM MM
Channel (MSCH) GMMREG GMMMM

GC NT DC

l Reception of MSCH is optional RRC
PDCP

PDCP
l MxCH channels are mapped on
Multicast Channel MCH UM
RLC
PUM

DCCH /
MCCH MTCH MSCH
DTCH

MAC
New SIB 13 informs MCH

about MBMS configuration PHY


MBSFN
MBSFN, Multicast/Broadcast Single Frequency Network

Cells belonging to MBSFN area are co-ordinated
and transmit a time-synchronized common waveform
eNodeBs within an MBSFN area are synchronized.
From point of view of the terminal, this appears to be a single
transmission as if originating from one large cell
(with correspondingly large delay spread).
Cyclic prefix is utilized to cover the difference in the propagation delays
from the multiple cells. MBMS therefore uses an extended cyclic prefix

MBSFN – MBMS Single Frequency Network

Mobile communication network Single Frequency Network
each eNode B sends individual each eNode B sends identical
signals signals


MBSFN

If network is synchronised,
Signals in downlink can be
combined


evolved Multimedia Broadcast Multicast Services
Multimedia Broadcast Single Frequency Network (MBSFN) area
l Useful if a significant number of users want to consume the same data
content at the same time in the same area!
l Same content is transmitted in a specific area is known as MBSFN area.
l Each MBSFN area has an own identity (mbsfn-AreaId 0…255) and can consists of
multiple cells; a cell can belong to more than one MBSFN area.
l MBSFN areas do not change dynamically over time.

MBSFN area 0 MBSFN area 255

11
3 8 13
MBSFN reserved cell.
1 6 12 15
A cell within the MBSFN
area, that does not support 4 9 14
MBMS transmission.
2 7 13
5 10
A cell can belong to
MBSFN area 1
more than one MBSFN
area; in total up to 8.


eMBMS
Downlink Channels
l Downlink channels related to MBMS
l MCCH Multicast Control Channel
l MTCH Multicast Traffic Channel
l MCH Multicast Channel
l PMCH Physical Multicast Channel

l MCH is transmitted over MBSFN in
specific subrames on physical layer

l MCH is a downlink only channel (no HARQ, no RLC repetitions)
l Higher Layer Forward Error Correction (see TS26.346)

l Different services (MTCHs and MCCH) can be multiplexed


eMBMS channel mapping

Subframes 0,4,5 and 9 are not MBMS, because
Of paging occasion can occur here

Subframes 0 and 5 are not MBMS, because
of PBCH and Sync Channels


eMBMS allocation based on SIB2 information

011010

Reminder:
Subframes
0,4,5, and 9
Are non-MBMS


eMBMS: MCCH position according to SIB13


LTE Release 9
Dual-layer beamforming
l 3GPP Rel-8 – Transmission Mode 7 = beamforming without
UE feedback, using UE-specific reference signal pattern,
l Estimate the position of the UE (Direction of Arrival, DoA),
l Pre-code digital baseband to direct beam at direction of arrival,
l BUT single-layer beamforming, only one codeword (TB),

l 3GPP Rel-9 – Transmission Mode 8 = beamforming with or
without UE feedback (PMI/RI) using UE-specific reference
signal pattern, but dual-layer,
l Mandatory for TDD, optional for FDD,
l 2 (new) reference signal pattern for two new antenna ports 7 and 8,
l New DCI format 2B to schedule transmission mode 8,
l Performance test in 3GPP TS 36.521 Part 1 (Rel-9) are adopted to
support testing of transmission mode 8.


LTE Release 9
Dual-layer beamforming – Reference Symbol Details

l Cell specific
antenna port 0 and
antenna port 1
reference symbols

Antenna Port 0 Antenna Port 1

l UE specific antenna
port 7 and antenna
port 8 reference
symbols

Antenna Port 7 Antenna Port 8


2 layer beamforming
throughput

Spatial multiplexing: increase throughput but less coverage

1 layer beamforming: increase coverage

SISO: coverage and throughput, no increase

2 layer beamforming
Increases throughput and
coverage

coverage
Spatial multiplexing increases throughput, but looses coverage


Location based services
l Location Based Services“
l Products and services which need location
information

l Future Trend:
Augmented Reality


Where is Waldo?


Location based services
The idea is not new, … so what to discuss?

Satellite based services

Location
controller

Network based services

Who will do the measurements? The UE or the network? = „assisted“

Who will do the calculation? The UE or the network? = „based“

So what is new?
Several ideas are defined and hybrid mode is possible as well,
Various methods can be combined.


E-UTRA supported positioning methods


LTE Release 9
LTE positioning
l The standard positioning methods supported for E-UTRAN
access are:
l network-assisted GNSS (Global Navigation Satellite System) methods
– These methods make use of UEs that are equipped with radio receivers
capable of receiving GNSS signals, e.g. GPS.
l downlink positioning
– The downlink (OTDOA – Observed Time Difference Of Arrival) positioning
method makes use of the measured timing of downlink signals received from
multiple eNode Bs at the UE. The UE measures the timing of the received
signals using assistance data received from the positioning server, and the
resulting measurements are used to locate the UE in relation to the
neighbouring eNode Bs.
l enhanced cell ID method
– In the Cell ID (CID) positioning method, the position of an UE is estimated with
the knowledge of its serving eNode B and cell. The information about the
serving eNode B and cell may be obtained by paging, tracking area update, or
other methods.


E-UTRA supported positioning network architecture
Control plane and user plane signaling
LCS4)
Client

S1-U Serving S5 Packet Lup
SUPL / LPP Gateway Gateway SLP1)
(S-GW) (P-GW) LCS
Server (LS) SLs
LPPa
LPP Mobile
Management E-SMLC2) GMLC3)
S1-MME SLs
Entity (MME)
LTE-capable device LTE base station
User Equipment, UE eNodeB (eNB)
(LCS Target)

Secure User Plane
Location positioning SUPL= user plane
protocol LPP = signaling
control plane 1) SLP – SUPL Location Platform, SUPL – Secure User Plane Location
2) E-SMLC – Evolved Serving Mobile Location Center
signaling 3) GMLC – Gateway Mobile Location Center
4) LCS – Location Service
5) 3GPP TS 36.455 LTE Positioning Protocol Annex (LPPa)
6) 3GPP TS 36.355 LTE Positioning Protocol (LPP)


E-UTRAN UE Positioning Architecture
l In contrast to GERAN and UTRAN, the E-UTRAN positioning
capabilities are intended to be forward compatible to other access
types (e.g. WLAN) and other positioning methods (e.g. RAT uplink
measurements).
l Supports user plane solutions, e.g. OMA SUPL 2.0

UE = User Equipment
SUPL* = Secure User Plane Location
OMA* = Open Mobile Alliance
SET = SUPL enabled terminal
SLP = SUPL locaiton platform
E-SMLC = Evolved Serving Mobile
Location Center
MME = Mobility Management Entity
RAT = Radio Access Technology

*www.openmobilealliance.org/technical/release_program/supl_v2_0.aspx

November 2012 | LTE Introduction | 171 Source: 3GPP TS 36.305

Global Navigation Satellite Systems

l GNSS – Global Navigation Satellite Systems; autonomous
systems: l GNSS are designed for
l GPS – USA 1995. continuous
l GLONASS – Russia, 2012. reception, outdoors.
l Gallileo – Europe, target 201?. l Challenging environments:
l Compass (Beidou) – China, under urban, indoors, changing
development, target 2015. locations.
l IRNSS – India, planning process.
E5a E1
L5 L5b L2 G2 E6 L1 G1
1164 1215 1237 1260 1300 1559 1591 f [MHz]
1563 1587 1610

GALLILEO GPS GLONASS
Signal fCarrier [MHz] Signal fCarrier [MHz] Signal fCarrier [MHz]
1602±k*0,562
E1 1575,420 L1C/A 1575,420 G1
5
1246±k*0,562
E6 1278,750 L1C 1575,420 G2
5
E5 1191,795 L2C 1227,600 k = -7 … 13
E5a 1176,450 L5 1176,450
http://www.hindawi.com/journals/ijno/2010/812945/
E5b 1207,140


Assisted GNSS (A-GNSS)

l The network assists the device GNSS receiver
to improve the performance in several aspects:
l Reduce GNSS start-up and acquisition times.
l Increase GNSS sensitivity, reduce power consumption.
l UE-assisted.
l Device (= User Equipment, UE) Source:
TS 36.355
transmits GNSS measurement LTE Positioning
Protocol (LPP)
results to network server, where
position calculation takes place.
l UE-based.
l UE performs GNSS measurements
and position calculation, supported by:
– Data to assist these measurements, e.g.
reference time, visible satellite list etc.
– Data providing for position calculation, e.g.
reference position, satellite ephemeris, etc.


LTE Positioning Protocol (LPP) 3GPP TS 36.355
LPP position methods
- A-GNSS Assisted Global Navigation Satellite System
- E-CID Enhanced Cell ID
LTE radio - OTDOA Observed time differerence of arrival
signal
*GNSS and LTE radio signals
eNB
Measurements based on reference sources*

Target LPP
Location
Device Server
Assistance data

LPP over RRC
UE Control plane solution E-SMLC Enhanced Serving
Mobile Location Center

LPP over SUPL
SUPL enabled
Terminal
SET User plane solution SLP SUPL location
platform


LPP and lower layers

LPP PDU Transfer
eNB


User plane stack
SUPL occupies the application layer in the LTE user plane stack,
with LPP (or another positioning protocol !) transported as another layer
above SUPL.


GNSS positioning methods supported
l Autonomeous GNSS

l Assisted GNSS (A-GNSS)
l The network assists the UE GNSS receiver to
improve the performance in several aspects:
– Reduce UE GNSS start-up and acquisition times
– Increase UE GNSS sensitivity
– Allow UE to consume less handset power
l UE Assisted
– UE transmits GNSS measurement results to E-SMLC where the position calculation
takes place
l UE Based
– UE performs GNSS measurements and position calculation, suppported by data …
– … assisting the measurements, e.g. with reference time, visible satellite list etc.
– … providing means for position calculation, e.g. reference position, satellite ephemeris, etc.


GNSS candidates and augmentation systems
l GPS/Modernized GPS – Global Positioning System (USA, since 1995)
l Galileo (Europe, under development, target 2013)
l GLONASS (Russia)
l Compass, Beidou-2 (China, under development, target 2015)
l IRNSS (India, in process of planning)

l SBAS – Satellite based augmentation systems
– Geostationary satellites supporting error corrections by overlay signals
– WAAS – Wide Area Augmentation System (USA)
– EGNOS – European Geostationary Navigation Overlay Service (Europa)
– MSAS – Mulit-Functional Satellite Augmentation System (Japan)
– QZSS – Quasi-Zenith Satellite System (Japan, supports GPS, expected 2013)
– GAGAN – GPS Aided Geo Augmented Navigation (India)


GNSS band allocations

E5a E1

L5 E5b L2 G2 E6 L1 G1 f/MHz
1164 1215 1237 1260 1300 1559 1563 1587 1591 1610


GPS and GLONASS satellite orbits

GPS:
26 Satellites
Orbital radius 26560 km

GLONASS:
26 Satellites
Orbital radius 25510 km


Position Determination
tsj trj
Rj = c·Δtj

l „Pseudo distance“ Satellite – Receiver
l Rj = c ·Δtj=c· (trj-tsj) = ρj + c·Δclock + ΔIono + ΔTropo + ΔMpath + ΔInt + ΔNoise
– ρj = real distance („error-free“)
– c·Δclock = Clock error (4th unknown variable → 4th satellite required)
– ΔIono , ΔTropo = „speed of light“ error due to ionosphere and troposhere
– ΔMpath , ΔInt , ΔNoise = trj uncertainty due to multipath propagation, interference,
noise

l Each satellite j broadcasts its current position ρSAT,j and local time tsj.
l With ρSAT,j and ρj the receiver position can be evalutated
l Additional signal phase measurements to increase accuracy


Backup basic terms

l Ephemeris
l Table of values that gives the positions of objects in the sky at a given time
l Almanach
l Set of data that every satellite transmits. It includes information about the
state (health) of the entire satellite constellation and coarse data on every
satellite‘s orbit.
l Cold start
l No ephemeris, almanac or location data available (reset state)
l Hot start
l Location data available


Why is GNSS not sufficent?

Critical scenario Very critical scenario GPS Satellites visibility (Urban)

l Global navigation satellite systems (GNSSs) have restricted
performance in certain environments

l Often less than four satellites visible: critical situation for GNSS
positioning
 support required (Assisted GNSS)
 alternative required (Mobile radio positioning)
Reference [DLR]


(A-)GNSS vs. mobile radio positioning
methods

(A-)GNSS Mobile radio systems
Low bandwidth (1-2 MHz) High bandwidth (up to 20 MHz for LTE)
Very weak received signals Comparatively strong received signals
One strong signal from the serving BS,
Similar received power levels from all satellites
strong interference situation
Long synchronization sequences Short synchronization sequences
Complete signal not a-priori known to
Signal a-priori known due to low data rates
support high data rates, only certain pilots
Very accurate synchronization of the satellites
Synchronization of the BSs not a-priori guaranteed
by atomic clocks
Line of sight (LOS) access as normal case Non line of sight (NLOS) access as normal case
 not suitable for urban / indoor areas  suitable for urban / indoor areas
3-dimensional positioning 2-dimensional positioning


Measurements for positioning

l UE-assisted measurements. l eNB-assisted measurements.
l Reference Signal Received l eNB Rx – Tx time difference.
Power l TADV – Timing Advance.
(RSRP) and Reference Signal – For positioning Type 1 is of
Received Quality (RSRQ). relevance.
l RSTD – Reference Signal Time l AoA – Angle of Arrival.
Difference. l UTDOA – Uplink Time Difference
l UE Rx–Tx time difference. of Arrival.
TADV (Timing Advance)
= eNB Rx-Tx time difference + UE Rx-Tx time difference
Neighbor cell j = (TeNB-RX – TeNB-TX) + (TUE-RX – TUE-TX)
UL radio frame #i
RSTD – Relative time difference
between a subframe received from
neighbor cell j and corresponding
subframe from serving cell i:
TSubframeRxj - TSubframeRxi DL radio frame #i UL radio frame #i
DL radio frame #i

Serving cell i
eNB Rx-Tx time difference is defined
UE Rx-Tx time difference is defined RSRP, RSRQ are as TeNB-RX – TeNB-TX, where TeNB-RX is the
as TUE-RX – TUE-TX, where TUE-RX is the measured on reference received timing of uplink radio frame #i
received timing of downlink radio frame signals of serving cell i and TeNB-TX the transmit timing of
#i from the serving cell i and TUE-TX the downlink radio frame #i.
transmit timing of uplink radio frame #i.

Source: see TS 36.214 Physical Layer measurements for detailed definitions


Observed Time difference

Observed Time
Difference of Arrival
OTDOA

If network is synchronised,
UE can measure time difference

Methods‘ overview
CID E-CID (RSRP/TOA/TADV) E-CID (RSRP/TOA/TADV) [Trilateration]

E-CID (AOA) [Triangulation] Downlink / Uplink (O/U-TDOA) [Multilateration] RF Pattern matching

To be updated!!


Cell ID

l Not new, other definition: Cell of Origin (COO).
l UE position is estimated with the knowledge of the geographical
coordinates of its serving eNB.
l Position accuracy = One whole cell .


Enhanced-Cell ID (E-CID)

l UE positioning compared to CID is specified more
accurately using additional UE and/or E UTRAN radio
measurements:
l E-CID with distance from serving eNB  position accuracy: a circle.
– Distance calculated by measuring RSRP / TOA / TADV (RTT).
l E-CID with distances from 3 eNB-s  position accuracy: a point.
– Distance calculated by measuring RSRP / TOA / TADV (RTT).
l E-CID with Angels of Arrival  position accuracy: a point.
– AOA are measured for at least 2, better 3 eNB‘s.

RSRP – Reference Signal Received Power
TOA – Time of Arrival
November 2012 | LTE Introduction | 189 TADV – Timing Advance
RTT – Round Trip Time

TADV – Timing Advance (Round Trip Time, RTT)
l Base station measures:
eNB Rx-Tx = TeNB-Rx – TeNB-Tx
eNB Rx Timing of subframe n

l eNB orders the device to
correct its uplink timing: TA = eNB Rx-Tx eNB Tx Timing of subframe n
eNB Rx-Tx.
UE Rx-Tx
l Timing Advance command UE Rx Timing of subframe n
(MAC).

UE Tx Timing of subframe n
l UE measures and reports:
UE Rx-Tx = TUE-Rx – TUE-Tx
l LPP_IE: ue-RxTxTimeDiff. l Distance of UE to eNB is
estimated as d=c*RTT/2.
l c = speed of light.
l eNB calculates and reports to
LS: TADV = eNB Rx-Tx + UE
Rx-Tx l Advantage: No
l LPPa_IE: timingAdvanceType1/2 synchronization
l TADV = Round Trip Time (RTT). between eNB‘s.


Angle of Arrival (AOA)

l AoA = Estimated angle of a UE with respect to a reference
direction (= geographical North), positive in a counter-
clockwise direction, as seen from an eNB.
l Determined at eNB antenna based
on a received UL signal (SRS).
l Measurement at eNB:
l eNB uses antenna array to estimate
direction i.e. Angle of Arrival (AOA).
l The larger the array, the more
accurate is the estimated AOA.
l eNB reports AOA to LS.
l Advantage: No synchronization
between eNB‘s.
l Drawback: costly antenna arrays.


OTDOA – Observed Time Difference of
Arrival
l UE position is estimated based on measuring TDOA of
Positioning Reference Signals (PRS) embedded into overall
DL signal received from different eNB’s.
l Each TDOA measurement describes a hyperbola (line of constant
difference 2a), the two focus points of which (F1, F2) are the two
measured eNB-s (PRS sources), and along which the UE may be
located.
l UE’s position = intersection of hyperbolas for at least 3 pairs of
eNB’s.


LTE Release 9
UE positioning – Reference Symbol Details

l PRS is a pseudo-random QPSK
sequence similar to CRS
l PRS pattern (baseline for further
discussion and CR drafting):
l Diagonal pattern with time
varying frequency shift,
l PRS mapped around CRS (to
avoid collisions)

Antenna Port 0


Positioning Reference Signals (PRS) for OTDOA
Definition

l Cell-specific reference signals (CRS) are not sufficient for
positioning, introduction of positioning reference signals
(PRS) for antenna port 6.
l SINR for synchronization
and reference signals of
neighboring cells needs to
be at least -6 dB.
l PRS is a pseudo-random
QPSK sequence similar
to CRS; PRS pattern:
l Diagonal pattern with time
varying frequency shift.
l PRS mapped around CRS to avoid collisions;
never overlaps with PDCCH; example shows
CRS mapping for usage of 4 antenna ports.


Uplink (UTDOA)

l UTDOA = Uplink Time Difference of Arrival
l UE positioning estimated based on:
– measuring TDOA of UL (SRS) signals received in different eNB-s
– each TDOA measurement describes a hyperbola (line of constant difference 2a),
the 2 focus points of which (F1, F2) are the two receiving eNB-s (SRS
receiptors), and along which the UE may be located.
– UE’s position = intersection of hyperbolas for at least three pairs of eNB-s
(= 3 eNB-s)
– knowledge of the geographical coordinates of the measured eNode Bs
l Method as such not specified for LTE  Similarity to 3G assumed

- eNB-s measure and
report to eNB_Rx-Tx to LS
-LS calculates UTDOA Location
Server
and estimates the UE
position


What is Self Organizing Networks, SON?
l SON = Self-Organizing Networks, methods for automatic configuration
and optimization of the network
l 3 components:
l Self-Configuration:

l basic setup: IP address configuration, association with a
Gateway, software download,…
l Initial radio configuration: neighbour list configuration,…
l Self-Optimization:
l Neighbour list optimization, coverage/capacity
optimization,…
l Self-Healing:

l Failure detection/localization,…


Motivation for SON

Heterogeneous Networks:

No way without SON

Source: Deutsche Telekom


Overview SON

Connect eNB

eNB failure


Architecture
Centralized SON
Distributed SON
Centralized
O&M SON

O&M O&M
SON SON

Hybrid SON

eNB eNB
SON SON


Energy Savings

l Match Network Capacitiy to the required traffic
l Switch on / off cell on demand

l Vodafone contribution on NGMN:


Self Healing
l Automatic detection of failures (Sleeping Cells)
l Usually detected by performance statistics
l Unreliable, because statistics sometimes fluctuate largely

l Cell Outage Compensation
l Recovery actions
– Fallback to previous SW
– Switching to backup units for HW
l Self optimization of the surrounding NW


Automatic Configuration of Physical Cell ID
Release 8

l Automatic configuration of new deployed eNodeBs
l 504 Physical Cell IDs are supported

l Selection of Physical Cell ID must be
l Collision free
– The ID is unique in the area the cell covers
l Confusion free
– A cell shall not have neighboring cells with identical IDs

l Definitely a must for Femto-Cells (HeNBs)

l Centralized or Distributed Architecture


Mobility Robustness Optimization (MRO)
Release 9

l Manually setting of HO parameters
l Time consuming → Often neglected
l Optimal settings may depend on momentary radio conditions
– Difficult to control manually

l Incorrect HO parameter setting may lead to
l Non-optimal use of network resources
– Unnecessary handovers
– Prolonged connection to a non-optimal cell
l HO failures
– Degradation of the service performance
l Radio Link failures
– Combined impact on user experience and network performance
– The main objective of MRO is to reduce RLF


Coverage and Capacity Optimization
Pilot pollution (Interference)
Cell edge performance
Trade-Off between coverage and capacity
 Coverage has priority

Measurement by Network:

eNodeB l Call drop rates
→ Indication of
eNodeB
insufficient coverage

l Traffic counters
eNodeB
→ Capacity problems

Detection of unintended coverage holes

Mobility Load Balancing (MLB) Optimization
Release 9


IMT-Advanced Requirements

l A high degree of commonality of functionality worldwide while
retaining the flexibility to support a wide range of services and
applications in a cost efficient manner,
l Compatibility of services within IMT and with fixed networks,
l Capability of interworking with other radio access systems,
l High quality mobile services,
l User equipment suitable for worldwide use,
l User-friendly applications, services and equipment,
l Worldwide roaming capability; and
l Enhanced peak data rates to support advanced services and
applications,
l 100 Mbit/s for high and
l 1 Gbit/s for low mobility
were established as targets for research,


Do you Remember?
Targets of ITU IMT-2000 Program (1998)
IMT-2000
The ITU vision of global wireless access
in the 21st century
Global
Satellite

Suburban Urban
In-Building

Picocell
Microcell
Macrocell

Basic Terminal
PDA Terminal
Audio/Visual Terminal

l Flexible and global
– Full coverage and mobility at 144 kbps .. 384 kbps
– Hot spot coverage with limited mobility at 2 Mbps
– Terrestrial based radio access technologies
l The IMT-2000 family of standards now supports four different multiple
access technologies:
l FDMA, TDMA, CDMA (WCDMA) and OFDMA (since 2007)


IMT Spectrum

MHz

MHz

Next possible spectrum
allocation at WRC 2015! MHz

MHz


Expected IMT-Advanced candidates

Long
Term
Evolution

Ultra
Mobile
Broadband

Advanced
Mobile
WiMAX

Source: TTA‘s workshop for the future of IMT-Advanced technologies, June 2008


IMT – International Mobile Communication

l IMT-2000
l Was the framework for the third Generation mobile communication
systems, i.e. 3GPP-UMTS and 3GPP2-C2K
l Focus was on high performance transmission schemes:
Link Level Efficiency
l Originally created to harmonize 3G mobile systems and to increase
opportunities for worldwide interoperability, the IMT-2000 family of
standards now supports four different access technologies, including
OFDMA (WiMAX), FDMA, TDMA and CDMA (WCDMA).
l IMT-Advanced
l Basis of (really) broadband mobile communication
l Focus on System Level Efficiency (e.g. cognitive network
systems)
l Vision 2010 – 2015


System level efficiency
l Todays mobile communication networks use static frequency allocation
l Network planning
l Adaptation to traffic load over cell boarder not possible
l Dynamic spectrum allocation to increase system efficiency
l Radio resource management to configure the occupied bandwidth of each cell
l Dynamic management for inter-cell interference reduction
l Cognitive networks (self organizing networks SON) which adjust automatically to
traffic load and interference conditions.
l Application oriented source and channel coding
l Typically, source and channel coding are separated, i.e. MPEG and convolutional
coding. Joint Source & Channel Coding (JSCC) promises better efficiency
l Base stations are on 24/7– but why?
l Basisstations/Relaystations operate in „sleep“ or „off“ modes
l Enhancements of interference situations and energy consumption
l Too many interfaces reduce the throughput
l Reducing the amount of components in the network structure
l Heterogenuous networks
l Usage of various radio access technologies of same core network


LTE-Advanced
Possible technology features

Relaying Wider bandwidth
technology support

Enhanced MIMO Cooperative
schemes for DL and UL base stations

Interference management Cognitive radio
methods methods

Radio network evolution Further enhanced
MBMS


Bandwidth extension with Carrier aggregation


LTE-Advanced
Carrier Aggregation
Component carrier CC

Contiguous carrier aggregation

Non-contiguous carrier aggregation

Aggregation
l Contiguous
l Intra-Band

l Non-Contiguous
l Intra (Single) -Band

l Inter (Multi) -Band

l Combination

l Up to 5 Rel-8 CC and 100 MHz
l Theoretically all CC-BW combinations possible (e.g. 5+10+20 etc)


Motivation

l 1) Higher data rate through
more spectrum allocation
l to fulfill 4G requirements
l 1 Gbit/s in downlink (up to 5
component carriers, up to 100
MHz)
l Other methods (spectral
efficiency etc.) already
exploited or not relevant

l 2) Exploit the total
(combined) BW assigned in
form of separated bands
l scattered spectrum


Carrier aggregation (CA)
General comments

l Two or more component carriers are aggregated in LTE-Advanced
in order to support wider bandwidths up to 100 MHz.
l Support for contiguous and non- Intra-band contiguous
Frequency band A Frequency band B
contiguous component carrier
aggregation (intra-band) and
inter-band carrier aggregation. Component
Carrier (CC)
l Different bandwidths per Intra-band non-contiguous
Frequency band A Frequency band B
component carrier (CC) are possible.
l Each CC limited to a max. of 110 RB
using the 3GPP Rel-8 numerology 2012 © by Rohde&Schwarz

(max. 5 carriers, 20 MHz each). Inter-band

l Motivation. Frequency band A Frequency band B

l Higher peak data rates to meet
IMT-Advanced requirements.
l NW operators: spectrum aggregation, enabling Heterogonous Networks.


Overview
l Carrier Aggregation (CA)
enables to aggregate up to 5 different
cells (component carriers CC), so that a
maximum system bandwidth of 100 MHz
can be supported (LTE-Advanced
requirement).
l Each CC = Rel-8 autonomous cell
Cell 1 Cell 2
– Backwards compatibility
l CC-Set is UE specific
– Registration  Primary (P)CC UE1 UE4 UE3 U3 UE4 U2

– Additional BW  Secondary (S)CC-s 1-4
l
CC2
Network perspective CC1

– Same single RLC-connection for one UE
(independent on the CC-s) UE1 UE2
– Many CC (starting at MAC scheduler)
CC2 CC1
UE3

operating the UE
l For TDD
– Same UL/DL configuration for all CC-s UE4


Carrier aggregation (CA)
General comments, cont’d.
l A device capable of carrier aggregation has 1 DL primary component
carrier and 1 associated primary UL component carrier.
l Basic linkage between DL and UL is signaled in SIB Type 2.
l Configuration of primary component carrier (PCC) is UE-specific.
– Downlink: cell search / selection, system information, measurement and
mobility.
– Uplink: access procedure on PCC, control information (PUCCH) on PCC.
– Network may decide to switch PCC for a device  handover procedure is
used.
l Device may have one or several secondary component carriers. Secondary
Component Carriers (SCC) added in RRC_CONNECTED mode only.
– Symmetric carrier aggregation.
– Asymmetric carrier aggregation (= Rel-10).
Downlink Uplink

SCC SCC SCC PCC SCC PCC SCC SCC SCC SCC

2012 © by Rohde&Schwarz

PDSCH and PDCCH PUSCH and PUCCH

PDSCH, PDCCH is optional PUSCH only


LTE-Advanced
Carrier Aggregation – Initial Deployment
l Initial LTE-Advanced deployments will likely be limited to
the use of two component carrier.
l The below are focus scenarios identified by 3GPP RAN4.


Bandwidth
 BWChannel(1)  BWChannel( 2)  0.1 BWChannel(1)  BWChannel( 2) 
Nominal channel spacing   0.3 MHz
l General 

0.6 


– Up to 5 CC
– Up to 100 RB-s pro CC
– Up to 500 RB-s aggregated

l Aggregated transmission
bandwidth
– Sum of aggregated channel
bandwidths
– Illustration for Intra band
contiguous
– Channel raster 300 kHz

l Bandwidth classes
– UE Capability


Bands / Band-Combinations (I)
l E-UTRA CA Band
l Band / Band-Combinatios specified in RAN4 for CA scenarios
l Already 4 CA Bands specified
– For inter frequency  bands without practical interest to guarantee quick
progress of the work


Bands / Band-Combinations (II)
l Under discussion
l 25 WI-s in RAN4 with
practical interest
– Inter band (1 UL CC)
– Intra band cont (1 UL CC)
– Intra band non cont
(1 UL CC)
– Inter band (2 UL CC)

l Release independency
l Band performance is
release independent
– Band introduced in Rel-11
– Performance tested for Rel-
10


Carrier aggregation - configurations
l CA Configurations
l E-UTRA CA Band + Allowed BW
= CA Configuration
– Intra band contiguous
– Most requirements
– Inter band
– Some requirements
– Main interest of many companies
– Intra band non contiguous
– No configuration / requirements
– Feature of later releases?

l CA Requirement applicability
– CL_X  Intra band CA
– CL_X-Y  Inter band
– Non-CA  no CA
(explicitely stated for the Test point
which are tested differently for CA
and not CA)


UE categories for Rel-10 NEW!
UE categories 6…8 (DL and UL)
Maximum number
Maximum number of bits Maximum number of
UE of DL-SCH transport Total number of
of a DL-SCH transport supported layers for
Category block bits received soft channel bits
block received within a TTI spatial multiplexing in DL
within a TTI
… … … … …
149776 (4 layers)
Category 6 301504 3654144 2 or 4
75376 (2 layers)
149776 (4 layers)
Category 7 301504 3654144 2 or 4
75376 (2 layers)
Category 8 2998560 299856 35982720 8

Maximum number
~3 Gbps peak Maximum number Support
of UL-SCH Total layer 2
DL data rate of bits of an UL-SCH for
UE transport buffer size
for 8x8 MIMO transport block 64Q
Category block bits [bytes]
transmitted within a AM
transmitted
TTI in UL
within a TTI
…
… … … …
3 300 000
Category 6 51024 51024 No
3 800 000
Category 7 102048 51024 No
42 200 000
Category 8 1497760 149776 Yes

~1.5 Gbps peak
November 2012 | LTE Introduction | 226 UL data rate, 4x4 MIMO

Deployment scenarios
3) Improve coverage

l #1: Contiguous frequency aggregation F1 F2

– Co-located & Same coverage
– Same f
l #2: Discontiguous frequency aggregation
– Co-located & Similar coverage
– Different f
l #3: Discontiguous frequency aggregation
– Co-Located & Different coverage
– Different f
– Antenna direction for CC2 to cover blank spots
l #4: Remote radio heads
– Not co-located
– Intelligence in central eNB, radio heads = only transmission
antennas
– Cover spots with more traffic
– Is the transmission of each radio head within the cell the
same?
l #5:Frequency-selective repeaters
– Combination #2 & #4
– Different f
– Extend the coverage of the 2nd CC with Relays


Physical channel arrangement in downlink

Each component
carrier transmits P- Each component
SCH and S-SCH, carrier transmits
Like Rel.8 PBCH,
Like Rel.8


LTE-Advanced
Carrier Aggregation – Scheduling
l There is one transport block (in Contiguous
Non-Contiguous spectrum allocation
absence of spatial multiplexing) RLC transmission buffer
and one HARQ entity per Dynamic
scheduled component carrier switching

(from the UE perspective),
l A UE may receive multiple Channel
coding
Channel
coding
Channel
coding
Channel
coding
component carriers
simultaneously, HARQ HARQ HARQ HARQ

l Two different approaches are
Data Data Data Data
discussed how to inform the UE mod. mod. mod. mod.
about the scheduling for each
band, Mapping Mapping Mapping Mapping

l Separate PDCCH for each carrier,
l Common PDCCH for multiple carrier, e.g. 20 MHz

[frequency in MHz]


LTE-Advanced
Carrier Aggregation – Scheduling
Non-Contiguous spectrum allocation
Contiguous
RLC transmission buffer
Dynamic
switching

Channel Channel Channel Channel
coding coding coding coding
Each component
HARQ HARQ HARQ HARQ
Carrier may use its
Data Data Data Data
mod. mod. mod. mod. own AMC,
Mapping Mapping Mapping Mapping = modulation + coding
e.g. 20 MHz
scheme

[frequency in MHz]


Carrier Aggregation – Architecture downlink
1 UE using carrier aggregation
Radio Bearers Radio Bearers

ROHC ... ROHC ROHC ... ROHC ROHC ROHC
PDCP ... PDCP
Security ... Security Security ... Security Security Security

Segm. Segm. Segm. Segm. Segm. Segm. Segm. Segm.
... ... ... RLC ...
RLC ARQ etc ARQ etc ARQ etc ARQ etc ARQ etc ARQ etc
CCCH BCCH PCCH CCCH

MCCH MTCH
Logical Channels Logical Channels

Unicast Scheduling / Priority Handling MBMS Scheduling Scheduling / Priority Handling

Multiplexing UE1 ... Multiplexing UEn Multiplexing Multiplexing
MAC MAC

HARQ ... HARQ
HARQ ... HARQ HARQ ... HARQ

Transport Channels
Transport Channels

BCH PCH MCH UL-SCH UL-SCH
DL-SCH DL-SCH DL-SCH DL-SCH on CC1 on CCz
on CC1 on CCx on CC1 on CCy

In case of CA, the multi-carrier nature of the physical layer is only exposed
to the MAC layer for which one HARQ entity is required per serving cell


Common or separate PDCCH per Component Carrier?

l No cross-carrier scheduling. up to 3 (4) symbols
per subframe 1 subframe = 1 ms
l PDCCH on a component carrier Time 1 slot = 0.5 ms

assigns PDSCH resources on the

Frequency
same component carrier (and

PDCCH

PDCCH
PUSCH resources on a single PDSCH PDSCH
linked UL component carrier).
l Reuse of Rel-8 PDCCH structure

PDCCH
(same coding, same CCE-based

PCC
PDSCH PDSCH

PDCCH
resource mapping) and DCI formats.
l Cross-carrier scheduling.
l PDCCH on a component carrier
can assign PDSCH or PUSCH PDSCH PDSCH

PDCCH

PDCCH
resources in one of multiple component
carriers using the carrier indicator field.
l Rel-8 DCI formats extended with
3 bit carrier indicator field. No cross-carrier Cross-carrier
l Reusing Rel-8 PDCCH structure (same scheduling scheduling
coding, same CCE-based resource
mapping).

Carrier aggregation: control signals + scheduling
Each CC has
its own control
channels,
like Rel.8

Femto cells:
Risk of interference!
-> main component
carrier will send
all control information.


Cross-carrier scheduling

l Main motivation for cross carrier scheduling: Interference
management for HetNet (eICIC); load balancing.
l Cross carrier scheduling is optional to a UE.
l Activated by RRC signaling, if not activated no CFI is present.
l Component carriers are numbered, Primary Component Carrier
(PCC) is always cell index 0.
l Scheduling on a component carrier is only possible from ONE
component, independent if cross-carrier scheduling is ON or OFF:
PCFICH PDCCH
PDCCH PDCCH
PDSCH start
signaled by RRC
PDSCH …
not possible,
transmission
Component Component can only be Component
Carrier #1 Carrier #2 scheduled by Carrier #5
one CC

l If cross carrier-scheduling active, UE needs to be informed about
PDSCH start on component carrier  RRC signaling.

DCI control elements: CIF field

New field: carrier indicator field gives information,
which component carrier is valid.
=> reminder: maximum 5 component carriers!

Carrier Indicator Field, CIF, 3bits

f f f

Component carrier 1 Component carrier 2 Component carrier N


PDSCH start field
Example: 1 Resource block
Of a component carrier
R

R

R
PCFICH PCFICH PCFICH

e.g. PCC e.g. SCC PDCCH
Secondary component
Primary component carrier
carrier PDSCH start
field indicates
position
In cross-carrier scheduling,the UE does not read the PCFICH
andPDCCH on SCC, thus it has to know the start of the
PDSCH

Component Carrier – RACH configuration
Downlink
Asymmetric carrier
Aggregation possible,
e.g. DL more CC than UL
Uplink

All CC use same RACH preamble
Each CC has its own RACH Network responds on all CCs

Only 1 CC contains RACH


Symmetry l Component Carrier
l DL-UL – However position of DC
not known to System
l Number of CC
Simulator
– TDD: DL = UL
– FDD: DL >= UL
l Bandwidth
l Both


Addition, modification, release of additional CC

l RRCConnectionReconfiguration Message contains new Rel-
10 information element:
DL: bandwidth, antennas, MBSFN subframe
configuration, PHICH configuration, PDSCH
configuration, TDD config (if TD-LTE SCell)
UL: bandwidth, carrier frequency, additional
spectrum emission, P-Max, power control info,
uplink channel configuration (PRACH, PUSCH)

UE-specifc information; DL: cross-carrier
scheduling, CSI-RS configuration, PDSCH
UL: PUSCH, uplink power control, CQI, SRS


Default EPS bearer setup (3GPP LTE Rel-8)
UE EUTRAN

Initial access and RRC connection establishment
attach request and PDN connectivity request

Additional information Authentication
being submitted by a
3GPP Rel-10 device… NAS security

UE capability procedure

AS security

RRC connection reconfiguration
Attach accept and default EPS bearer context request

Default EPS bearer context accept


UE capability information transfer
3GPP Rel-10 add on’s Not all combinations
are allowed!

l New IE’s within the for a 3GPP Rel-10 device:
Carrier aggregation capabilities
are signaled separately for DL, UL.

What frequency bands, band combination (CA
and MIMO capabilities, inter-band, intra-band
contiguous and non-contiguous, bandwidth class
does the device support?

2012 © by Rohde&Schwarz

!


Carrier Aggregation - Activation

l A new MAC control element for Component Carrier Management is
defined containing at least the activation respectively deactivation
command for the secondary DL component carriers configured for
a UE. The new MAC CE is identified by a unique LCID
l
l For actual deactivation and activation signalling for the DL SCCs,
the MAC CE for CC Management includes a 4/5-bit bitmap where
each bit is representing one of the DL CCs that can be configured
in the UE. A bit set to 1 denotes activation of the corresponding DL
CC, a bit set to 0 respectively denotes deactivation

l New timer for implicit CC deactivation

l CC's are "just" additional resources. UL scheduling will assume
we do not have different QOS (delay/loss) on different CC's


Carrier Aggregation

l The transmission mode is not constrained to be the same
on all CCs scheduled for a UE

l A single UE-specific UL CC is configured semi-statically for
carrying PUCCH A/N, SR, and periodic CSI from a UE

l Frequency hopping is not supported simultaneously with
non-contiguous PUSCH resource allocation

l UCI cannot be carried on more than one PUSCH in a given
subframe.


Carrier Aggregation

l Working assumption is confirmed that a single set of PHICH
resources is shared by all UEs (Rel-8 to Rel-10)

l If simultaneous PUCCH+PUSCH is configured and there is
at least one PUSCH transmission
l UCI can be transmitted on either PUCCH or PUSCH with a
dependency on the situation that needs to be further discussed
l All UCI mapped onto PUSCH in a given subframe gets mapped onto
a single CC irrespective of the number of PUSCH CCs


UE Architectures

l Possible TX architectures
l Same / Different antenna (connectors) for each CC
l D1/D2 could be switched to support CA or UL MIMO


LTE–Advanced solutions from R&S
R&S® SMU200 Vector Signal Generator


LTE–Advanced solutions from R&S
R&S® FSQ Signal Analyzer
RBW 2 MHz
VBW 5 MHz
Ref -20 dBm Att 5 dB SWT 2.5 ms

-20

A
-30

1 AP
CLRWR
-40

-10 dB
-50

-60

20 MHz 20 MHz
E-UTRA carrier 2
-70
E-UTRA carrier 2 EXT

fc,E-UTRA carrier 2 = 2135 MHz fc,E-UTRA carrier 2 = 2205 MHz3DB
-80

-90

-100

-110

-120

Center 2.17 GHz 10 MHz/ Span 100 MHz

LTE-Advanced – An introduction
Date: 8.OCT.2009 14:13:24 Roessler | October 2009 | 247
A.

Enhanced MIMO schemes

l Increased number of layers:
l Up to 8x8 MIMO in downlink.
l Up to 4x4 MIMO in uplink.

l In addition the downlink reference signal structure has been
enhanced compared with LTE Release 8 by:
l Demodulation Reference signals (DM-RS) targeting PDSCH demodulation.
– UE specific, i.e. an extension to multiple layers of the concept of Release 8 UE-
specific reference signals used for beamforming.
l Reference signals targeting channel state information (CSI-RS) estimation
for CQI/PMI/RI/etc reporting when needed.
– Cell specific, sparse in the frequency and time domain and punctured into the data
region of normal subframes.


Downlink reference signals in LTE-Advanced

l Define two types of RS, l RS and data are subject to the same
l RS targeting PDSCH demodulation, pre-coding operation,
l RS targeting CSI generation (for l complementary use of Rel-8 CRS by
CQI/PMI/RI/etc reporting when the UE is not precluded,
needed), l RS targeting CSI generation (for
l RS targeting PDSCH demodulation LTE-A operation) are
(for LTE-Advanced operation) are l Cell specific and sparse in frequency
l UE specific and time
– Transmitted only in scheduled RBs and l Rel-8 transmission schemes using
the corresponding layers Rel-8 cell-specific and/or UE-
– Different layers can target the same or specific RS still supported,
different UEs
– Design principle is an extension of the
concept of Rel-8 UE-specific RS (used for
beamforming) to multiple layersDetails on
UE-specific RS pattern, location, etc are
FFS
l RS on different layers are mutually
orthogonal,


Cell specific Reference Signals vs. DM-RS
LTE Rel.8 LTE-Advanced (Rel.10)
CRS0 DM-RS0 CRS0 + CSI-RS0

s1 s1

s2 s2
Pre- Pre-
........

........
coding coding

sN sN

Cell specific DM-RSN Cell specific
Reference signalsN reference signalsN +
Channel status information
reference signals 0

l Demodulation-Reference signals DM-RS and data are precoded
the same way, enabling non-codebook based precoding and
enhanced multi-user beamforming.

Ref. signal mapping: Rel.8 vs. LTE-Advanced
l Example:
LTE (Release 8) LTE-A (Release 10) l 2 antenna ports, antenna port 0,
0 1 2 3 4 5 6 0 1 2 3 4 5 6 0 1 2 3 4 5 6 0 1 2 3 4 5 6 CSI-RS configuration 8.
l PDCCH (control) allocated in the

1 0
x x x x
first 2 OFDM symbols.

R E L E A S E
x x x x l CRS sent on all RBs; DTX sent
for the CRS of 2nd antenna
x x x x
port.
x x x x
l DM-RS sent only for scheduled

8
x x x x
RBs on all antennas; each set
R E L E A S E
x x x x coded differently between the
two layers.
x x x x

x x x x l CSI-RS punctures Rel. 8 data;
sent periodically over allotted
PDSCH PDCCH CRS DM-RS CSI-RS REs (not more than twice per
frame)


DL MIMO
Extension up to 8x8
Codeword to layer mapping for spatial multiplexing

l Max number of transport blocks: 2 Number
Number Codeword-to-layer mapping
l Number of MCS fields of layers
of code
words i  0,1, M symb  1
layer

l one for each transport block
x ( 0) (i)  d ( 0) (2i)
l ACK/NACK feedback x (1) (i)  d ( 0) (2i  1)
l 1 bit per transport block for evaluation 5 2 M symb  M symb 2  M symb 3
layer ( 0) (1)

x (i)  d (3i )
( 2) (1)
as a baseline x (3) (i )  d (1) (3i  1)

l Closed-loop precoding supported x ( 4) (i)  d (1) (3i  2)

l Rely on precoded dedicated x ( 0) (i )  d ( 0) (3i)
x (1) (i )  d ( 0) (3i  1)
demodulation RS (decision on DL RS) x ( 2) (i)  d ( 0) (3i  2)

l M symb  M symb 3  M symb 3
layer ( 0) (1)
6 2
Conclusion on the codeword-to-
x (3) (i)  d (1) (3i)
layer mapping: x ( 4) (i)  d (1) (3i  1)
x (5) (i )  d (1) (3i  2)
l DL spatial multiplexing of up to eight
x ( 0) (i )  d ( 0) (3i)
layers is considered for LTE-Advanced, x (1) (i )  d ( 0) (3i  1)
x ( 2) (i)  d ( 0) (3i  2)
l Up to 4 layers, reuse LTE codeword-to-
7 2 M symb  M symb 3  M symb 4
layer ( 0) (1)

layer mapping, x (3) (i )  d (1) (4i )
x ( 4) (i )  d (1) (4i  1)
l Above 4 layers mapping – see table x (5) (i )  d (1) (4i  2)
x ( 6) (i )  d (1) (4i  3)
l Discussion on control signaling x ( 0) (i )  d ( 0) (4i )
x (1) (i )  d ( 0) (4i  1)
details ongoing x ( 2) (i )  d ( 0) (4i  2)
x (3) (i )  d ( 0) (4i  3)
layer ( 0) (1)

x ( 4) (i)  d (1) (4i)
x (5) (i)  d (1) (4i  1)
x ( 6) (i )  d (1) (4i  2)
x ( 7 ) (i)  d (1) (4i  3)

MIMO – layer and codeword
101
Codeword 1: 101
111000011101 Codeword 1:
100 111000011101
011
ACK
000

Codeword 2: 111 Codeword 2:
010101010011 010101010011
001
111
Up to 8 times ACK
the data -> Receiver only
8 layers Sends
2 ACK/NACKs

Scheduling of Transmission Mode 9 (TM9)
l NEW DCI format 2C with 3GPP Rel-10.
l Used to schedule transmission mode 9 (TM9), which is spatial
multiplexing with DM-RS support of up to 8 layers (multi-layer
transmission).
– DM-RS scrambling and number of layers are jointly signaled in a 3-bit
field.
l DCI format 2C.
– Carrier indicator [3 bit]. One Codeword:
Codeword 0 enabled,
Two Codewords:
Codeword 0 enabled,
– Resource allocation header [1 bit] Codeword 1 disabled Codeword 1 enabled
Value Message Value Message
– Resource Allocation Type 0 and 1.
0 1 layer, port 7, n =0 0 2 layers, ports 7-8, nSCID=0
– TPC command for PUCCH [2 bit]. SCID

1 1 layer, port 7, nSCID=1 1 2 layers, ports 7-8, nSCID=1
– Downlink Assignment Index1) [2 bit]. 2 1 layer, port 8, nSCID=0 2 3 layers, ports 7-9
– HARQ process number 3 1 layer, port 8, nSCID=1 3 4 layers, ports 7-10
[3 bit (FDD), 4 bit (TDD)]. 4 2 layers, ports 7-8 4 5 layers, ports 7-11

– Antenna ports, scrambling identiy 5 3 layers, ports 7-9 5 6 layers, ports 7-12

and # of layers; see table [3 bit]. 6 4 layers, ports 7-10 6 7 layers, ports 7-13

– SRS request1) [0-1 bit]. 7 Reserved 7 8 layers, ports 7-14

– MCS, new data indicator, RV for 2 transport block [each 5 bit]. 1) TDD only


Uplink MIMO
Extension up to 4x4
l Rel-8 LTE.
l UEs must have 2 antennas for reception.
l But only 1 amplifier for transmission is available (costs/complexity).
l UL MIMO only as antenna switching mode (switched diversity).

l 4x4 UL SU-MIMO is needed to fulfill peak data rate requirement of
15 bps/Hz.

l Schemes are very similar to DL MIMO modes.
l UL spatial multiplexing of up to 4 layers is considered for LTE-Advanced.
l SRS enables link and SU-MIMO adaptation.

l Number of receive antennas are receiver-implementation
specific.
l At least two receive antennas is assumed on the terminal side.


UL MIMO – signal generation in uplink
Similar to Rel.8 Downlink:

codewords layers antenna ports

Modulation Transform Resource element SC-FDMA
Scrambling mapper
Layer
Precoding
mapper
Modulation Transform Resource element SC-FDMA
Scrambling mapper

Avoid
periodic bit QPSK Up to DFT,
sequence 16-QAM 4 layer as in Rel 8,
64-QAM but non-
contiguous
allocation possible


UL MIMO – layers and codewords
Number of layers Number of codewords Codeword-to-layer mapping
i  0,1,...,M symb  1
layer

1 1 x (0) (i)  d (0) (i) M symb  M symb
layer ( 0)

x (0) (i)  d (0) (2i )
2 1 x (1) (i)  d (0) (2i  1) M symb  M symb 2
layer ( 0)

x (0) (i)  d (0) (i)
2 2 M symb  M symb  M symb
layer ( 0) (1)

x (i)  d (i)
(1) (1)

x (0) (i)  d (0) (i)

3 2 x (1) (i)  d (1) (2i ) M symb  M symb  M symb 2
layer ( 0) (1)

x ( 2) (i)  d (1) (2i  1)

x (0) (i)  d (0) (2i )
x (1) (i)  d (0) (2i  1)
layer ( 0) (1)

x ( 2)
(i)  d (1)
(2i )
x (3)
(i)  d (1)
(2i  1)
Up to 4
Layers 2 codewords
(Rel.11)


UL MIMO scheduling – DCI format 4 NEW!
l Carrier indicator [0-3 bit]. l Downlink Assignment Index [2
bit].
l Resource Block Assignment:
l TDD only, UL-DL config. 1-6.
l [bits] for Resource Allocation
Type 0 l CSI request [1 or 2 bit].
l



 log ( N UL ( N UL  1) / 2) 
2 RB RB 

 2 bit for cells with more than two
cells in the DL (carrier
aggregation).
l [bits] for Resource Allocation
Type 1. l SRS request [2 bit].
l Resource Allocation Type [1
   N RB / P  1 
UL

log 2  

 

bit].

  4  
 l Transport Block 1.
l TPC command for PUSCH [2 l MCS, RV [5 bit].
bit]. l New data indicator [1 bit].
l Cyclic shift for DM RS and l Transport Block 2.
OCC index [2 bit]. l MCS, RV [5 bit].
l UL index [2 bit]. l New data indicator [1 bit].
l TDD only for UL-DL l Precoding information.
configuration 0.


LTE-Advanced
Enhanced uplink SC-FDMA
l The uplink
transmission
scheme remains
SC-FDMA.
l The transmission of
the physical uplink
shared channel
(PUSCH) uses DFT
precoding.
l Two enhancements:
l Control-data
decoupling
l Non-contiguous
data transmission


Physical channel arrangement - uplink

Simultaneous transmission of
Clustered-DFTS-OFDM
PUCCH and PUSCH from
= Clustered DFT spread
the same UE is supported
OFDM.

Non-contiguous resource block allocation => will cause higher
Crest factor at UE side


LTE-Advanced

Unused subcarriers


LTE-Advanced

Due to distribution we get active subcarriers beside
non-active subcarriers: worse peak to average ratio,
e.g. Crest factor


Simultaneous PUSCH-PUCCH transmission, multi-
cluster transmission
l Remember, only one UL carrier in 3GPP Release 10;
scenarios:
l Feature support is indicated by PhyLayerParameters-v1020 IE*).
PUCCH and PUCCH and fully
allocated PUSCH allocated PUSCH

f [MHz] f [MHz]
PUCCH PUSCH
partially PUCCH and partially
allocated PUSCH allocated PUSCH

f [MHz] f [MHz]
*) see 3GPP TS 36.331 RRC Protocol Specification


Benefit of localized or distributed mode
„static UE“: frequency selectivity is not time variant -> localized allocation

Multipath causes frequency
Selective channel,
It can be time variant or
Non-time variant

„high velocity UE“: frequency selectivity is time variant -> distributed allocation


Resource Allocation types in the uplink

l Uplink Resource Allocation Type 0.
l Contiguous allocation as today in 3GPP Release 8.

l Uplink Resource Allocation Type 1.
l UL bandwidth is divided into two sets of Resource Blocks (RB).
l Each set has a number of Resource Block Groups (RBG) of size P.
l Combinatorial index r, indicates RBG starting and ending index for
both set of RB (s0, s1-1 | s2, s3-1. System RBG
Bandwidth Size (P)
M 1 N  si
si i 0
M 1
r 
i 0 M i
1  si  N , si  si 1 ≤10
11 – 26
1
2

– M = 4, N is bandwidth dependent, N  N RB / P  UL
 1 27 – 63
64 – 110
3
4


Multi-cluster allocation
Uplink Resource Allocation Type 1
l Example: LTE 10 MHz (50 RB), P = 3 leads to 17 RBG.
l Combinatorial index r indicates RBG starting indices for RB set #1
(s0, s1-1, defining cluster #1) and RB set #2 (s2, s3-1, defining cluster
#2).
l Range for s0, s1, s2, s3 for 10 MHz (50 RB): 1 to 18 (see previous
slide).
l Applied rule: s0 < s1 < s2 < s3 (see previous slide).
l Example: s0=2, s1=9, s2=10, s3=11:
“START” “END” “START”
(s0) (s1-1) (s2) “END”
(s3-1)

RBG#1 RBG#3 RBG#5 RBG#7 RBG#9 RBG#11 RBG#13 RBG#15

RBG#0 RBG#2 RBG#4 RBG#6 RBG#8 RBG#10 RBG#12 RBG#14 RBG#16

Cluster #1 Cluster #2


What are the effects of “Enhanced SC-FDMA”?

Source: http://www.3gpp.org/ftp/tsg_ran/WG4_Radio/TSGR4_54/Documents/R4-101056.zip


Significant step towards 4G: Relaying ?



Radio Relaying approach

No Improvement of SNR resp. CINR


L1/L2 Relaying approach



LTE-Advanced
Relaying
l LTE-Advanced extends LTE Release 8 with support for
relaying in order to enhance coverage and capacity
l Classification of relays based on
l implemented protocol knowledge…
– Layer 1 (repeater)
– Higher Layer (decode and forward or even mobility management, session
set-up, handover)
l … and whether the relay has its own cell identity
– Type 1 relay effectively creates its own cell (own ID and own
synchronization and reference channels
– Type 2 relay will not have its own Cell_ID Type 2

Type 1


LTE-Advanced: Relaying

The relay node (RN) is wirelessly connected to a donor cell of a
donor eNB via the Un interface, and UEs connect to the RN via
the Uu interface

Uu Un EPC

UE RN eNB

Inband: 2 links operate in the same frequency
Outband: 2 links operate in different frequencies


Co-operative Relaying Approaches

l Receiver in MS (or BS) gains from both, the signal origin and the relay
station
l Co-operative Relaying creates virtual MIMO

• RSs are „decode-and-forward“ devices


LTE-Advanced
Coordinated Multipoint Tx/Rx (CoMP)

CoMP

Coordination between cells

Coordinated Multipoint, CoMP
Controller
Controller UE estimates
Various DL
CSI
+ feedback feedback
Info about CoMP,
UE perform signal
processing
Controller
CoMP
Each eNB Info UE estimates various
Estimates Uplink Downlink + feedback
CSI No info about CoMP
feedback
at UE


LTE-A: PCFICH indicating PDCCH size
PCFICH content in LTE R-8 PCFICH content in LTE-A
Subframe where Number of OFDM symbols for PDCCH when

PCFICH is sent N RB  10
DL N RB  10
DL PCFICH can
Subframe 1 and 6 1, 2 2 indicate up to 4
in TDD mode OFDM symbols
Subframe 0 in FDD 1, 2, 3 2, 3, 4
mode
for used by PDCCH
PDSCH PBCH
PDCCH S-SCH
PCFICH P-SCH

Time
Frequency


LTE-A: Multiple Access MAC layer concept
Transport block 1 Transport block K

Segmentation Segmentation

FEC FEC FEC FEC

Non-frequency adaptive Frequency adaptive
Resource scheduler
Packet processing
Mapping on dispersed chunks Mapping on frequency optimal chunks

Bit interleaved coded modulation Bit interleaved coded modulation
Quasi cyclic block low density partity check Quasi cyclic block low density partity check

Antenna summation, linear precoding

IFFT + CP insertion

RF generation

LTE-Advanced: C-Plane latency

Connected
10ms
Dormant Active
Already fullfills ITU
50ms Requirement with Rel.8,
Idle but ideas to speed up:

•Combined RRC Connection Request and NAS Service Request
•Reduced processing delays in network nodes
•Reduced RACH scheduling period: 10ms to 5ms

Shorter PUCCH cycle: requests are sent faster
Contention based uplink: UE sends data without previous request


LTE Registration – R8 to R10…. UE SS
incl. security activation
RRC ConnectionnRequest
RRC ConnectionnSetup
RRC ConnectionSetupComplete

contains
NAS ATTACH REQUEST

NAS PDN CONNECTIVITY REQUEST

NAS : AUTHENTICATION REQUEST

Registration NAS : AUTHENTICATION RESPONSE

Already in RRC NAS : SECURITY MODE COMMAND

Connection NAS : SECURITY MODE COMMAND COMPLETE

Request RRC : SECURITY MODE COMMAND
RRC : SECURITY MODE COMMAND COMPLETE

RRC ConnectionReconfiguration
NAS ATTACH ACCEPT
contains
NAS : ACTIVATE DEFAULT EPS BEARER CONTEXT REQ

RRC ConnectionReconfigurationComplete
NAS : ATTACH COMPLETE

NAS : ACTIVATE DEFAULT EPS BEARER CONTEXT
ACCEPT


Present Thrust- Spectrum Efficiency
Momentary snapshot of frequency spectrum allocation

Why not use this
part of the spectrum?
l FCC Measurements:- Temporal and geographical variations in the utilization of the assigned
spectrum range from 15% to 85%.


ODMA – some ideas…
BTS

Mobile devices behave as
relay station


Cooperative communication
How to implement antenna arrays in mobile handsets?

Multi-access

Independent
fading paths

Each mobile
is user and relay 3 principles of aid:
•Amplify and forward
•Decode and forward
•Coded cooperation

Cooperative communication

Multi-access

Independent
fading paths

Virtual Higher signaling
Antenna System complexity
Array Cell edge coverage
Group mobility

Mobile going GREEN


Ubiquitous communication


There will be enough topics
for future trainings 

Thank you for your attention!

Comments and questions
welcome!


LTE Evolution: From Release 8 to Release 10

More Related Content

What's hot (20)

Viewers also liked (6)

Similar to LTE Evolution: From Release 8 to Release 10 (20)

More from Rohde & Schwarz North America (20)

Recently uploaded (20)

LTE Evolution: From Release 8 to Release 10