Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Space Diversity
Spatial Multiplexing
Beamforming
LTE/MIMO 표준기술 2
Introduction
SIMO and MISO
SIMO
A single transmit antenna and NR receive antennas
Receive Spatial (Antenna) Diversity
MISO
NT transmit antennas and a single receive antenna
Transmit Spatial (Antenna) Diversity
LTE/MIMO 표준기술 4
MIMO
NT transmit antennas and NR receive antennas
Diversity gain ~ NTNR
Capacity ~ min(NT, NR)
Performance-oriented MIMO
Data-rate-oriented MIMO
LTE/MIMO 표준기술 5
Space (Antenna) Diversity
Receive Antenna Diversity
Selection Diversity
Antenna element with the highest SNR (or other metric) is selected
Greatest SNR improvement when
Desired signal subject to independent (uncorrelated) fading and signal receiver with the same
average power at each element
Background noise is AWGN, equal power and uncorrelated across elements
Equal Gain Combining (EGC)
Co-phasing and summing the decision statistics at each element
LOS signal (no fading): equivalent to finding a beamforming weight vector w
Rayleigh fading signals: no longer optimum, but tends to achieve more diversity than
selection diversity
Maximal Ratio Combining (MRC)
Each sensor output is co-phased and weighted by its SNR before combining
MRC require design in receiver circuitry to achieve correct weighting factors
Optimal in the sense that provides best statistical reduction of fading of any linear combiner
Shows little improvement over EGC for Gaussian noise channel
LTE/MIMO 표준기술 7
Rx MRC and Tx STC
So S1*
antenna
*
|H(f)| S1 S0
… …
f
LTE/MIMO 표준기술 9
BER of MRC & STC
huge gain
x1 h0 n1 | ho |2 | h1 |2
x1
x h s n SNR
s
TX
s ho RX
2 1 2 2
(MRC)
x2
h1 x hs n
LTE/MIMO 표준기술 12
Spatial Multiplexing
MIMO Spatial Multiplexing (SM)
Multiple Input Multiple Output (MIMO)
Multiple antennas at both transmitter and receiver
MIMO uses multipath to advantage to “multiply data rate”
Transmits different data along different paths (simplified view)
MxN MIMO can multiply data rate by M or N (whichever is less) if there is enough
multipath.
Best in urban high-multipath environment (and indoors)
Less effective in suburban and rural low-multipath environments
LTE/MIMO 표준기술 14
How can we get multiplexing?
Simple concept
(1) Transmit “one” data in one link (1 Tx & 1 Rx antenna)
(2) Transmit “two” data in two links far away from each other (1 Tx & 1 Rx antenna, respectively)
(3) Transmit “two” data in one link (1 Tx & 1 Rx antenna) ??
(4) Transmit “two” data in one link (2 Tx & 2 Rx antenna) ??
LTE/MIMO 표준기술 15
MIMO Data Rate Exercise
Example system configuration
System BW = 100 MHz
Highest Modulation = 64QAM
Channel Coding = {1/3, 1/2} + {repetition, puncturing}
MIMO order = 8 x 8
LTE/MIMO 표준기술 16
SVD MIMO as a closed-loop MIMO
In CL-SU-MIMO, SVD-MIMO is the optimum
LTE/MIMO 표준기술 17
MIMO Channel Decomposition
LTE/MIMO 표준기술 18
MIMO Channel Decomposition
1 ~
w1
x y
~
~
x V VH ~
w
U UH y
n min
nmin
Channel
Pre-processing Post-processing
y Hx w
x C nt , y C nr , w ~ Ν (0, N 0I nr )
LTE/MIMO 표준기술 19
Channel Diagonalization
~
yU y
H
U (Hx w )
H
U H (UDV H x w )
U H (UDV H V~x w)
D~
x UHw
~
y D~ ~
xw
LTE/MIMO 표준기술 20
SVD Example (Mathematics)
H UDV H
0 1 0 0 0
1 0 0 0 2 0 0 4 0 0
0
0 1 0 0
0 0 3 0 0 0 1 0 0 0 3 0 0 0
0 0 1 0
0.2 0 0 0 0.8
0 0 0 0 0 0 0 0 1 0 0 5 0 0
0 0 0 1 0
0 4 0 0 0 1 0 0 0 0 0 0 0 0
0.8 0 0 0 0.2
UU I
H
0 0 1 0 0 0 0 1 1 0 0 0
0 1 0 0 0 1 0 0 0 1 0 0
0 0 0 1 1 0 0 0 0 0 1 0
1 0 0 0 0 0 1 0 0 0 0 1
VV H I
0 0 0.2 0 0.8 0 1 0 0 0 1 0 0 0 0
1 0 0 0 0 0 0 1 0 0 0 1 0 0 0
0 1 0 0 0 0.2 0 0 0 0.8 0 0 1 0 0
0 0 0 1 0 0 0 0 1 0 0 0 0 1 0
0 0.2 0.8 0 0 0 0.2 0 0 0 0 1
0 0.8 0
0 1 0 0 0
* singular value decomposition is not unique. 0 0 1 0 0
VH 0.2 0 0 0 0.8
0.4 0 0 0.5 0.1
LTE/MIMO 표준기술 21
0.4 0 0 0.5 0.1
Singular Values
The Singular Values of the channel matrix are the key to determining
how good the channel is for MIMO
SISO: C BW log(1 )
2
MIMO: C
i 1...N
BW log
1
N
i ( H )
LTE/MIMO 표준기술 22
SU-MIMO & MU-MIMO
Single-user MIMO schemes
User 1
Data 1
Data 2 BS Channel 2
Larger diversity gain than single- User 2
Data 3
available
LTE/MIMO 표준기술 23
MU-MIMO in theory
Unlike SU-MIMO which is the point-to-point multiple-antenna channels, MU-MIMO is
the multiple-antenna broadcast channels.
The multiuser capacity of multiple-antenna broadcast channels depends heavily on
whether the transmitter knows the channel coefficients to each user.
For example, in a Gaussian broadcast channel with M transmit antennas and n single-antenna
users, the sum rate capacity scales like M log log n for large n if perfect channel state
information (CSI) is available at the transmitter, yet only logarithmically with M if it is not.
The MU-MIMO with perfect CSI at the transmitter is well known capacity achieving
dirty paper coding (DPC) [1].
On the other hand, other techniques such as random beamforming (RBF) [2] and
opportunistic beamforming [3] have been studied as the sub-optimum scheme where
there is only partial or very little side information available at the transmitter.
[1] M. H. M. Costa, ''Writing on dirty paper,'' IEEE Trans. Inform. Theory, vol.IT-29, no.3, May 1983
[2] M. Sharif and B. Hassibi, "On the capacity of MIMO broadcast channels with partial side
information," IEEE Trans. Inform. Theory, vol.51, no.2, Feb. 2005
[3] P. Viswanath, D. N. C. Tse, and R. Laroia, "Opportunistic beamforming using dumb antennas,"
IEEE Trans. Inform. Theory, vol.48, no.6, June 2002
LTE/MIMO 표준기술 24
MU-MIMO
Allocate each antenna resource to different user
Scheduler selects a user with the highest rate per antenna
Increase system throughput!
Find the unitary matrix which gives the max sum rate
Feedback the unitary matrix and the associated SINRs
LTE/MIMO 표준기술 29
Multi-BS MIMO
COMP (CO-operative/Co-ordinated MultiPoint transmission and
reception)
16m & LTE-Advanced
Soft handover
ICIC
LTE/MIMO 표준기술 30
Beamforming
BF Starting Point
interference
cells
interference
user user
LTE/MIMO 표준기술 32
Beamforming & AAS
Several small wide-beamwidth antennas form large directional array with narrow
beamwidth
Higher antenna gain improves range
Narrow beam reduces interference to and from other sectors and cells
Nulls in antenna pattern can be directed toward interferers
Improves channel capacity & spectral efficiency
Base station only (Tx and Rx)
LTE/MIMO 표준기술 33
Multi Antenna System
Spatial Diversity
Gain due to multiple fading paths
Reduces fading margin
Gain depends upon scattering environment
Coherent Gain
Improves link budget
Focuses radiation = passive I.C.
Gain = 10*log(N_ant)
Interference Mitigation
Active interference cancellation
Enhances capacity
Increases per-user data rate
Spatial Multiplexing
Single user = enhanced throughput
Multiple users = SDMA
LTE/MIMO 표준기술 34
CL-MU-MIMO and SDMA?
“OL-SU-MIMO SM” (where different antennas clearly transmit
independent different symbols, i.e., symbols are multiplexed in
physical antenna domain) and “SDMA” (where different antennas
work together to make spatially separate multiple signatures for
multiple symbol delivery) look different from each other
However, “CL-MU-MIMO” and “SDMA” can be regarded as the same
MU-MIMO „typically‟ with codebook-based precoding/beamforming
SDMA „typically‟ with non-codebook-based precoding/beamforming
LTE/MIMO 표준기술 35
MIMO Application in LTE
R8 LTE MIMO Design Principles
Anti-fading: TD (and SM)
Enhance spectral efficiency: SM (TD and BF)
Enhance SNR: BF (and codebook-based precoding)
Channel adaptive: CL precoding, Rank adaptation
LTE/MIMO 표준기술 37
R8 LTE MIMO Summary
Cell-specific transmit diversity scheme
One scheme for all control channels except synchronization signals
Transmit (transparent TD (e.g. PVS, TSTD) can be used for synchronization
Diversity signals)
2 Tx ant: SFBC
4 Tx ant: SFBC+FSTD
DL Open loop and Closed loop
SU-MIMO
Up to 2 codewords transmission
Codebook-based precoding
MU-MIMO
Single layer per UE
LTE/MIMO 표준기술 38
SFBC+FSTD
When the pilot density is same for data transmitted
on four antennas
LTE/MIMO 표준기술 39
DL Physical Channel Processing
code words layers antenna ports
LTE/MIMO 표준기술 40
Basic Structure of LTE MIMO
LTE/MIMO 표준기술 41
Layer Mapping
The complex-valued modulation symbols for each of the code words
to be transmitted are mapped onto one or several layers.
Complex-valued modulation symbols (q)
d ( q) (0),...,d ( q) (M symb 1) for code
word q shall be mapped onto the layers
x(i) x (0) (i) ... x ( 1) (i) ,
T
i 0,1,...,M symb
layer
1.
LTE/MIMO 표준기술 42
Layer Mapping for SM
LTE/MIMO 표준기술 43
Layer Mapping for Tx Diversity
LTE/MIMO 표준기술 44
Precoding
The precoder takes as input a block of vectors
x(i) x (0) (i) ... x ( 1) (i) ,
T
i 0,1,...,M symb
layer
1 from the layer mapping and generates a block of
vectors , i 0,1,...,M
y(i) ... y ( p) (i) ...
T ap
symb 1 to be mapped onto resources
on each of the antenna ports where y ( p) (i) represents the signal for
antenna port p .
LTE/MIMO 표준기술 45
Precoding without CDD
y (0) (i ) x ( 0) (i )
W (i )
y ( P 1)
(i ) x ( 1 )
(i )
LTE/MIMO 표준기술 46
Precoding for large-delay CDD
y (0) (i ) x (0) (i )
W (i ) D(i )U
y ( P 1) (i) x ( 1) (i )
LTE/MIMO 표준기술 47
Codebook for Precoding – 2 ports
For transmission on two antenna ports, p 0,1, the precoding matrix W (i) shall be
selected from Table 6.3.4.2.3-1 or a subset thereof.
LTE/MIMO 표준기술 48
Codebook for Precoding – 4 ports
The quantity Wn{s}denotes the matrix defined by the columns given by the set {s} from
the expression Wn I 2ununH unH un where I is the 4x4 identity matrix and the vector u n is
given by Table 6.3.4.2.3-2.
LTE/MIMO 표준기술 49
Codebook for Transmit Diversity
y (0) (2i ) 1 0
j 0 Re x (0) (i )
(1)
y (2i ) 1 0 1 0 j Re x (i )
(1)
y (0) (2i 1)
(1)
2 0 1
0 j Im x (0) (i )
y (2i 1)
1 0 j 0 Im x (i )
(1)
y (0) (4i ) 1 0 0 0 j 0 0 0
(1) 0 0
y (4i ) 0 0 0 0 0 0
y ( 2) (4i ) 0 1 0 0 0 j 0 0
( 3)
y (4i ) 0 0 0 0 0 0 0 0
y ( 0) (4i 1)
0 1
0 0 0 j
0 0 Re x ( 0) (i )
y (1) (4i 1) 0 0 0 0 0 0
0 0 Re x (1) (i )
( 2)
y (4i 1)
1 0
0 0 j 0
0 0 Re x ( 2) (i )
y (3) (4i 1)
1 0 0 0 0 0 0
0 0 Re x (3) (i )
( 0)
y (4i 2)
2 0 0 0 0 0 0
0 0 Im x ( 0) (i )
y (1) (4i 2)
( 2)
0 0 1 0 0 0
j 0 Im x (1) (i )
y (4i 2)
0 0 0 0 0 0
0 0 Im x ( 2) (i )
( 3)
y (4i 2)
0 0
0 1 0 0 0
j Im x (3) (i )
y ( 0) (4i 3) 0 0 0 0 0 0 0 0
(1) 0 0
y (4i 3) 0 1 0 0 0 j
y ( 2) (4i 3) 0 0 0 0 0 0 0 0
( 3)
y (4i 3) 0 0 1 0 0 0 j 0
LTE/MIMO 표준기술 50
RI, PMI, CQI
Rank Indication (RI)
Info about the channel rank or, expressed differently, the number of layers that
should, preferably, be used for Dl transmission to the UE
Precoder Matrix Indication (PMI)
Provides a precoder matrix that should, preferably, used for DL transmission
The reported precoder matrix should be determined assuming the number of layers
indicated by RI
PMI is reported only in CL mode. In OL mode, the network instead selects the
precoder matrix to use according to a pre-defined rule.
Network has 2 choices:
The network may follow the recommendation from UE, in which case it only has to confirm
with 1 bit indicator in DL scheduling assignment.
The network may select a different precoder. Then the precoder information needs to be
explicitly included in DL scheduling assignment.
Channel Quality Indication (CQI)
A set of pre-defined MCS combinations
LTE/MIMO 표준기술 51
Channel-Status Reports
Illustration of PMI Selection, CQI and RI Reporting (example)
LTE/MIMO 표준기술 52
Periodic/Aperiodic Channel Info Feedback
Periodic reporting Aperiodic reporting
When to send Periodically every 2-160 ms When requested by eNB
LTE/MIMO 표준기술 53
Sounding RS
Reference for channel quality information
Channel quality measurement for frequency/time aware scheduling
Channel quality measurement for link adaptation
Channel quality measurement for power control
Channel quality measurement for MIMO, BF
Sequence generation
r SRS n ru(,v ) n which is defined by cyclic shift of a base sequence ru ,v (n)
LTE/MIMO 표준기술 54
7 Tx modes are possible in PDSCH
Mode 1: Single-antenna port; port 0
DL transmissions using a single Tx antenna at eNB
LTE/MIMO 표준기술 59
LTE Performance with MIMO
Spectral Efficiency per Features
The percentage figure to the left represents the individual feature impact, and the percentage figure to the
right represents the accumulated impact of the features combined.
It is seen that in both downlink and uplink, the normalized FDD and TDD performance is indeed similar.
The differences are due to the TDD guard period, the UpPTS, and the slightly longer channel-quality-
feedback delays for TDD
* David Astély, et al, “LTE: The Evolution of Mobile Broadband,” IEEE Commun. Mag. April 2009
LTE/MIMO 표준기술 61
LTE DL Efficiency Benefit over R6 HSDPA
in Macro Cells*
LTE Benefit Gain Explanation
HSDPA suffers from intra-cell interference fro
Up to +70%
the Rake receiver. Rake receiver is assumed
OFDM with freq. depending on
in R6. However, most HSDPA terminals have
domain EQ the multi-path
an EQ that removes most intra-cell
profile
interference.
Frequency domain scheduling is possible in
Freq. domain packet OFDM system, but not in single carrier
+40%
scheduling HSDPA. The dual carrier HSDPA can get part
of the frequency domain scheduling gain.
No MIMO defined in HSDPA R6. The gain is
MIMO +15% relative to single antenna BS transmission.
HSDPA R7 includes MIMO.
Inter-cell interference The interference rejection combining works
+10%
rejection combining better in OFDM system with long symbols.
Total =3.0x 1.7 x 1.4 x 1.15 x 1.1
* Harri Holma and Antti Toskala, LTE for UMTS: OFDMA and SC-FDMA Based Radio Access, Wiley, 2009
LTE/MIMO 표준기술 62
Comparison with WiMAX MIMO
MIMO Mode/Capability Comparison w/ LTE
* Qinghua표준기술
LTE/MIMO Li, et. al., “MIMO Techniques in WiMAX and LTE: A Feature Overview”, IEEE Commun. Mag., May 2010 65
Implementation Issues
Power Penalty in MIMO Transmission
Transmission power divided among multiple antennas
2x2 MIMO is okay
but, 8x8 MIMO is…?
실제 구현 시에는 어떻게 각 안테나 element당 max power를 준비할까?
Single antenna: 20W/antenna-element
2x2 MIMO: 10W/antenna-element?
4x4 MIMO or 4x1 BF: 5W/antenna-element ?
Antenna selection, antenna grouping을 하는 경우는?
LTE/MIMO 표준기술 67
MIMO 성능: SINR
MIMO는 high SINR에서 SIMO대비 성능 이득을 보임
G-factor = SINR = S / (I+N)
LTE/MIMO 표준기술 70
Correlation with MIMO Repeater
Isolation : Tx path와 Rx path간의 isolation 발진문제
Coupling: Tx path간 또는 Rx path간의 coupling correlation
* 삼성전자
LTE/MIMO 표준기술 71
MIMO in Handset?
Okay w/ 2x2
What about 8x8, 12x12?
R0 R0
R0 R0
R0 R0
l 0 l 6 l 0 l 6
R0 R0 R1 R1
Two antenna ports
R0 R0 R1 R1
Not used for transmission on this antenan port
R0 R0 R1 R1
Reference symbols on this antenna port
R0 R0 R1 R1
l 0 l 6 l 0 l 6 l 0 l 6 l 0 l 6
R0 R0 R1 R1 R2 R3 R3
Four antenna ports
R0 R0 R1 R1 R2
R0 R0 R1 R1 R2 R3 R3
R0 R0 R1 R1 R2
l 0 l 6 l 0 l 6 l 0 l 6 l 0 l 6 l 0 l 6 l 0 l 6 l 0 l 6 l 0 l 6
LTE/MIMO 표준기술
even-numbered slots odd-numbered slots even-numbered slots odd-numbered slots even-numbered slots odd-numbered slots even-numbered slots odd-numbered slots
73
Antenna port 0 Antenna port 1 Antenna port 2 Antenna port 3
RI, PMI, CQI
Rank Indication (RI)
Info about the channel rank or, expressed differently, the number of layers that
should, preferably, be used for Dl transmission to the UE
Precoder Matrix Indication (PMI)
Provides a precoder matrix that should, preferably, used for DL transmission
The reported precoder matrix should be determined assuming the number of layers
indicated by RI
PMI is reported only in CL mode. In OL mode, the network instead selects the
precoder matrix to use according to a pre-defined rule.
Channel Quality Indication (CQI)
A set of pre-defined MCS combinations
Wideband or subband?
얼마나 정확하게? (e.g. direct channel reporting)
얼마나 주파수 축 상으로 촘촘하게? (subband case)
얼마나 자주?
LTE/MIMO 표준기술 74
CQI Compression Schemes
Wideband feedback
Best-M average (UE-selected sub-band feedback)
Higher layer-configured sub-band feedback
Additionally, delta compression can be used in combination with above options
LTE/MIMO 표준기술 75
Demodulator Complexity
ZF (Zero Forcing)
Brute-force : r Hx n, ~
x H 1r H 1n
Noise enhancement
MMSE (Minimum Mean Square Error)
Consider noise, too
A lot of sub-optimal MLD
MLD
Optimal performance, but high complexity
~x1
r1 h11 h12
2
~ arg min 송신기가 보낼 수있는 모든 { x1 , x 2 } 조합
x2 { x1 , x 2 } r2
h21 h22
Sub-Optimal MLD
ZF MMSE MLD
성능
Interference
Optimum Linear
Cancellation
Zero forcing
Maximum likelihood
Decision feedback minimum mean
Equalization (ISI) sequence detection
equalization (DFE) square error
(MLSD)
(MMSE)
Successive/parallel
Optimum multiuser interference Decorrelating,
Multiuser
detector (MUD) cancellation, MMSE
iterative MUD
ML detector sphere Bell Labs Layered
Spatial-multiplexing Zero forcing,
decoder (near Spaced Time
Receiver MMSE
optimum) (BLAST)
LTE/MIMO 표준기술 77
Summary
Why MIMO SM?
SISO
MISO Diversity
SIMO Diversity
MIMO Diversity
MIMO SM
LTE/MIMO 표준기술 79