[PPT] - Interference Alignment and RAMCOORAN Co-Ordinated Multi-Point with PowerPoint Presentation

SLIDE 1

1

RAMCOORAN

Per Zetterberg

Interference Alignment and Co-Ordinated Multi-Point with 802.11ac-feedback: Testbed Results

SLIDE 2

2

RAMCOORAN

Astonishing Result

K-transmitters and K-receivers, K-links: K/2 simultaneous interference-free links. Requires coding over multiple channel realizations. Global channel knowledge required.

Cadambe/Jafar, ”Interference Alignment and Degrees of Freedom of the K-User Interference Channel”, IEEE Trans, Information Theory 2008.

SLIDE 3

3

RAMCOORAN

Channel extension (Cadambe/ Jafar)

[ ]

( ) ( )

( )

m m ij ij ij

t f h t f h diag H , , , , 1

1

 =

Channel extended MIMO channel (my wording)

5 . 1 1 2 1 3 1 2 => + + + = n n n m

SLIDE 4

4

RAMCOORAN

The ”alignment”

( )

D N D

x w x x x w y

H H

= + + + = ...

1

SLIDE 5

5

RAMCOORAN

Implementation IA

BS 1 BS 2 BS 3 MS 1 MS 2 MS 3 Feedback: Wired ethernet 𝒘1 𝒘2 𝒘3 𝒗1 𝒗2 𝒗3

SLIDE 6

6

RAMCOORAN

Implementation: CoMP

BS 1 BS 2 BS 3 MS 1 MS 2 MS 3 Feedback: Wired ethernet 𝒘1 𝒘2 𝒘3 𝒗1 𝒗2 𝒗3

SLIDE 7

7

RAMCOORAN

Beamformer

“Approaching the Capacity of Wireless Networks through Distributed Interference Alignment", by Krishna Gomadam, Viveck R. Cadambe and Syed A. Jafar.

∑ ≠

+ =

k n n n k k k k k k k 2 2 , * 2 , *

~ σ v H u v H u SNIR

        =

∑

2 , 2 2

001 . , max ~

n n k N

H σ σ

SLIDE 8

8

RAMCOORAN

Beamformer initialization CoMP

[ ]

H 1 1 1 13 , 12 11 1

, V S U H H H H = =

[ ]

H 2 2 2 23 , 22 21 2

, V S U H H H H = =

[ ]

H 3 3 3 33 , 32 31 3

, V S U H H H H = =

( ) ( ) ( )

[ ]

1 :, 1 :, 1 :, ~

3 2 1

V V V V =

( )

1

~ ~ ~

−

= V V V W

H

SLIDE 9

9

RAMCOORAN

The testbed

3BS 10m 10m 3MS P=10dBm NF=10-11dB

SLIDE 10

10

RAMCOORAN

IEEE 802.11ac feedback

Training symbols One per antenna

IEEE 802.11ac feedback

SLIDE 11

11

RAMCOORAN

H

U S V H =

Number of phis and psis : NcNr-Nc^2/-Nc/2 Range 0, Range Quantization bits =7 or 9 Quantization bits 5 or 7

( ) [ ]

( )

c r r i r N i i

N N N i l l i T l i N r N c i j j i i

I G e e D V

× + = − = −

      =

∏ ∏

~ 1 , , , 1

1 1 , m i n 1 1

, . 1 . .

ψ

φ φ



The whole 6x2 matrix is treated by the three MSs.

π 2

2 / , 0 π

IEEE 802.11ac feedback, contd.

SLIDE 12

12

RAMCOORAN

Frequency granularity

Our implementation:

14MHz 1

38

19, -18, -17, -16, -15, -14, -13, -12, -11, -10, -9, -8,-6, -5, -4, -3, -2, -1, 1, 2, 3, 4, 5, 6, 8, 9,

10, 11,12, 13, 14, 15, 16, 17, 18, 19

2

20

18, -16, -14, -12, -10, -8, -6, -4, -2, -1, 1, 2, 4, 6,8, 10, 12, 14, 16, 18

4

8

12, -8, -4, -1, 1, 4,8, 12

8

6

16, -8, -1, 1, 8, 16

16

4

16,-1,1,16

SLIDE 13

13

RAMCOORAN

SNR feedback

14MHz 1

20

18, -16, -14, -12, -10, -8, -6, -4, -2, -1, 1, 2, 4, 6,8, 10, 12,14, 16, 18";

2

10

16, -12, -8, -4, -1, 1, 4, 8, 12, 16

4

8 12, -8, -4, -1, 1, 4, 8,12

8

4

8, -1, 1, 8

16

2

16,16

SLIDE 14

14

RAMCOORAN

Interpolate between subcarriers

Reuse V on adjacent subcarriers. Interpolate SNR values Reconstruct H as: H=diag(SNR0,SNR1)*V.

SLIDE 15

15

RAMCOORAN

Overhead

Based on: Improved MU-MIMO Performance for Future 802.11 Systems Using Differential Feedback, Ron Porat

The three short frames ≈ 2*60µs+num_users*60µs
Gaps=2 SIFS + num_users SIFS
Number of feedback bits per V matrix= (NcNr-Nc^2/-Nc)

(7+9)

Number of feedback for SNR num_users*8*SNRs*Nr

Feedback time in our case≈ 350µs+1065µs/(Ng*R) Ng=4, R=2 => 500µs With update interval : 20ms => Overhead 2.5%

SLIDE 16

16

RAMCOORAN

P P 1 P 2 P 3 P 4 P 5 Training frame Payload frames.. MCS0-9 20ms MCS0-9 MCS0-9

1. BS1 sends P0,P1
2. BS2 sends P2,P3
3. BS3 sends P4,P5
4. MS1-MS3 sends compressed V matrix to BS1.
5. BS1 un-compresses calculates all beamformes

using max SINR on 38subcarriers.

6. Each BS sends beamformed frames.
7. MSs saves all data.
8. Post-processing of received signals.

Our implementation

SLIDE 17

17

RAMCOORAN

Measurement environment

B0 B2 B1

C C

SLIDE 18

18

RAMCOORAN

Results LoS (stationary)

1 2 4 8 16 38 2 4 6 8 10 12 14 16 Ng

Sum througput bits/symbol/subcarrier Measured sum throughput

CoMP IA TDMA MIMO full-reuse SIMO full-reuse MIMO

SLIDE 19

19

RAMCOORAN

Results NLoS (stationary)

1 2 4 8 16 38 2 4 6 8 10 12 14 16 Ng Sum througput bits/symbol/subcarrier Measured sum throughput CoMP IA TDMA MIMO full-reuse SIMO full-reuse MIMO

SLIDE 20

20

RAMCOORAN

Why is NLoS better for IA?

10
5

5 10 15 20

10
5

5 10 15 20 25 C/Imax (dB) C/Imin (dB)

LoS NLoS

SLIDE 21

21

RAMCOORAN

Time varying channels

SLIDE 22

22

RAMCOORAN

5 10 15 1 2 3 4 5 6

Frame number Throughput

Time varying channels

Minute before People walking Minute after

SLIDE 23

23

RAMCOORAN

Time varying channels

5 10 15 1 2 3 4 5 6

Frame number Throughput

Simulated on measured channels

SLIDE 24

24

RAMCOORAN

Time varying channels

5 10 15 1 2 3 4 5 6

Frame number Throughput

0.5ms feedback delay

SLIDE 25

25

RAMCOORAN

Conclusion

IA gives 25% sum throughput gain over SU MIMO
n stationary channel (LoS and NLoS averaged)
CoMP gives 71% gain over SU MIMO.
All schemes limited by RF impairments.
Geometry factor imortant for IA versus MIMO.
Full reuse SIMO and MIMO worse than SU MIMO.
Next step: analyzing measurements with time

varying channels.

SLIDE 26

26

RAMCOORAN

Results versus EVM model

Schem e Actual perform ance I m pairm ent m odel IA 11.1 11.7 CoMP 15.2 17.3 SU MIMO 8.9 8.1 FR SIMO 6.5 6.2 FR SIMO 2.3 2.3