SLIDE 1
2
- TSC. Grupo Comunicaciones
GCOM-UC3M
Introduction and overview Non-coherent transmission schemes for Massive MIMO Multi-user DMPSK-based massive MIMO
- SINR and Error probability
- Channel coding to reduce the number of antennas
massive but feasible Ana Garca Armada Communications Research Group - - PowerPoint PPT Presentation
massive but feasible Ana Garca Armada Communications Research Group - - PowerPoint PPT Presentation
Non-coherent Large Scale MIMO: massive but feasible Ana Garca Armada Communications Research Group (GCOM) Universidad Carlos III de Madrid, Spain GCOM-UC3M Agenda Introduction and overview Non-coherent transmission schemes for
SLIDE 2
SLIDE 3
3
- TSC. Grupo Comunicaciones
GCOM-UC3M
New requirements call for new technologies
Introduction
Non coherent processing may be a good solution if combined with massive MIMO
SLIDE 4
4
- TSC. Grupo Comunicaciones
GCOM-UC3M
3 dB loss of non-coherent (NC) vs coherent (C) processing When we consider the needs of channel state information
(CSI) obtaining and sharing, this loss may become negligible
- A. Goldsmith’s work: To train or not to train? Channel
estimation is wasteful in some circumstances (channels with low coherence time, low SNR)
NC massive MIMO: the perfect match!
- The “magic” of massive MIMO (self interference cancellation)
may improve NC performance
- CSI estimation and sharing is vey complex in massive MIMO
(pilot contamination ...)
Non coherent communications – why now?
SLIDE 5
5
- TSC. Grupo Comunicaciones
GCOM-UC3M
Benefits of increasing (a lot) the number of antennas
- Improve data rates and reliability (multiplexing and diversity
gains)
- Decrease required transmit power
- Very simple precoders/decoders
Most usual configuration is
Massive MIMO
… R antennas at BS, R >> K single antenna users, K<<R … MU- massive MISO MU- massive SIMO
SLIDE 6
6
- TSC. Grupo Comunicaciones
GCOM-UC3M
[1] proposed an Uplink ASK (amplitude shift keying) energy-detector scheme
- For a single user, they achieve rates which are not different from coherent
schemes in a scaling law sense
- 2-user system proposed in [2]
- Too many antennas are required for reasonable performance with actual
constellation design.
[3] proposed decision-feedback differential detection of DMPSK. Relies on
particular channel model and similarities to IR-UWB
- Assumes the users to be randomly distributed in front of a large linear antenna
array at the BS. Not general.
Non-coherent massive MIMO
[1] M. Chowdhury, A. Manolakos, A.J. Goldsmith, “Design and Performance of Noncoherent Massive SIMO Systems,” Proc. of 48th Annual Conference on Information Sciences and Systems, Princeton, 2014. [2] M. Chowdhury, A. Manolakos, A.J. Goldsmith, “CSI is not needed for Optimal Scaling in Multiuser Massive SIMO Systems,” Proceedings of ISIT., Honolulu, July 2014. [3] A. Schenk, R.F.H. Fischer, “Noncoherent Detection in Massive MIMO Systems,” Proc. of International ITG/IEEE Workshop on Smart Antennas, Stuttgart, March 2013.
SLIDE 7
7
- TSC. Grupo Comunicaciones
GCOM-UC3M
One base station (BS) with R receive antennas K Mobile Stations (MSs) with single antenna Data symbol sequences sj[n] (j=1,…K) are M-PSK:
|sj,m[n]|= 1
Tx signal at time instant n comes from differentially encoding sj[n] : No channel coding (by now)
Multi-user Large Scale single input-multiple output (SIMO) uplink [4]
[4] A. G. Armada and L. Hanzo, “A Non-Coherent Multi-User Large Scale SIMO System Relying on M-ary DPSK,” IEEE ICC, Jun. 2015 pp 2517-2522.
… … MU- massive SIMO
SLIDE 8
8
- TSC. Grupo Comunicaciones
GCOM-UC3M
System model
x1[n] xK[n]
… …
y1[n] yR[n] y1[n-1] * y1[n] yR[n-1] * yR[n]
… S
z[n]
+
n1[n]
+
nR[n]
+
Hb
At the Rx: the phase difference of two consecutive symbols received at each antenna is non-coherently detected and they are all added to give the decision variable z[n] Reference SNR
SLIDE 9
9
- TSC. Grupo Comunicaciones
GCOM-UC3M
We define the joint symbol as
can be estimated from z[n]
Joint constellation:
Multiple users - detection
- 2
- 1
1 2
- 2
- 1
1 2
- 2
- 1
1 2
- 2
- 1
1 2
- 2
- 1
1 2
- 2
- 1
1 2
Asymptotic behavior: From the Law of Large Numbers we know that So we have
SLIDE 10
10
- TSC. Grupo Comunicaciones
GCOM-UC3M
Signal to Interference plus Noise Ratio (SINR)
- 20
- 15
- 10
- 5
5 10 15 20
- 30
- 20
- 10
10 20
reference SNR dB SINR dB
b2=8 b2=1
energy efficiency scaling with R, same as with perfect CSI R=100
SLIDE 11
11
- TSC. Grupo Comunicaciones
GCOM-UC3M
Performance – 2 users, DQPSK, SNR=0 dB
10
1
10
2
10
3
10
4
10
- 6
10
- 5
10
- 4
10
- 3
10
- 2
10
- 1
10
R Pe simul [REF] upper [REF] Design A Design B
[5] M. Chowdhury, A. Manolakos, A.J. Goldsmith, “CSI is not needed for Optimal Scaling in Multiuser Massive SIMO Systems,” Proceedings of ISIT., Honolulu, July 2014. [5]
SLIDE 12
12
- TSC. Grupo Comunicaciones
GCOM-UC3M
Performance – 2 users (M-ary PSK)
5 10 15 20 25 10
- 10
10
- 8
10
- 6
10
- 4
10
- 2
10
SINR dB Pe DQPSK simul DQPSK union bound DQPSK lower bound DBPSK simul DBPSK union bound DBPSK lower bound 8-DPSK simul 8-DPSK union bound 8-DPSK lower bound
DBPSK DQPSK 8-DPSK
SLIDE 13
13
- TSC. Grupo Comunicaciones
GCOM-UC3M
Design B – multiuser, M=4, SNR=0 dB
10
2
10
3
10
4
10
- 10
10
- 8
10
- 6
10
- 4
10
- 2
10
R Pe 1 user UB 1 user LB 2 users UB 2 users LB 3 users UB 3 users LB 4 users UB 4 users LB
SLIDE 14
14
- TSC. Grupo Comunicaciones
GCOM-UC3M
Comparison with coherent MRC – 2 dB loss
10
1
10
2
10
3
10
- 4
10
- 3
10
- 2
10
- 1
10
R Pe Design B, =0 dB 8-PSK MRC, =0dB QPSK MRC, =0dB Design B, =2 dB
CSI is estimated with a realistic error, which is also assumed to be Gaussian rate-loss of 33% due to pilot overhead
- > 8-PSK vs QPSK
SLIDE 15
15
- TSC. Grupo Comunicaciones
GCOM-UC3M
Channel coding and iterative decoding
RSC: recursive systematic convolutional URC: unity-rate code BCJR: Bahl-Cocke-Jelinek-Raviv algorithm
SLIDE 16
16
- TSC. Grupo Comunicaciones
GCOM-UC3M
BER improvement SNR=0dB
SLIDE 17
17
- TSC. Grupo Comunicaciones
Comparison with [6], SNR=0dB
GCOM-UC3M
[6] M. Chowdhury, A. Manolakos and Andrea Goldsmith,“Scaling Laws for Noncoherent Energy-Based Communications in the SIMO MAC,” IEEE Transaction on Information Theory, vol. 62, no. 4, Apr. 2016
[6]
SLIDE 18
18
- TSC. Grupo Comunicaciones
Number of antennas vs coding rate
GCOM-UC3M
SLIDE 19
19
- TSC. Grupo Comunicaciones
GCOM-UC3M
DMPSK for massive MIMO does not need CSI Improved performance wrt previous work Not far from coherent systems when CSI is noisy and pilot
- verhead is taken into account
Coding reduces the number of antennas to feasible values
Conclusions
SLIDE 20
20
- TSC. Grupo Comunicaciones