Random Access Protocols for Massive MIMO

Size: px

Start display at page:

Download "Random Access Protocols for Massive MIMO"

Stanley Watkins
5 years ago
Views:

Larsson Linköping University Sweden 2016 Tyrrhenian International Workshop

1 Random Access Protocols for Massive MIMO Elisabeth de Carvalho Jesper H. Sørensen Petar Popovski Aalborg University Denmark Emil Björnson Erik G. Larsson Linköping University Sweden 2016 Tyrrhenian International Workshop on Digital Communications (TIW16) Sept 12-14, 2016 Department of Electronic Systems

2 CSI acquisition and data transmission in Crowd Scenarios Megacities Machine type communications Hotspots Department of Electronic Systems

3 Presentation Content Massive MIMO: massive number of spatial degrees of freedom Crowds: exploit very large multiplexing gain Uplink Training based on orthogonal pilot sequences Length/number of orthogonal pilots limited Pilot shortage One solution: random access to the pilots and possibly the data 3

4 CSI Acquisition in Massive MIMO Time-division duplexing and channel reciprocity CSI is acquired using uplink training Exploited for downlink transmission Orthogonal pilots 4

5 Orthogonal Pilots are a limited resource Number of orthogonal pilots = pilot sequence length Pilot Sequence length limited by: # of pilots limited by channel coherence time # of pilots limited by transmit power Pilot Shortage Number of pilot sequences is much smaller than the number of pilot sequences 5

6 Traffic Burstiness Non-streaming internet applications Video streaming Social network Cloud time Machine-type communications: crowd of devices transmitting sporadically unpredictable and intermittent traffic 6

7 Random Pilot Access Crowd of devices with unpredictable and intermittent traffic Makes pilot pre-allocation very inefficient Need for: Scalable and efficient pilot access and data transmission protocols. 7 Proposed solution: Random Access to Pilot sequences

8 Random Access to Pilots Total number of terminals K and τ p orthogonal pilots available Users select a pilot sequence uniformly at random with probability p a Random pilot selection COLLISIONS Collisions = Pilot contamination Intra-Cell Pilot contamination 8

9 Pilot Collision UL Pilot transmission Users with same pilot sequence: their channel cannot be distinguished: g = all colliders g i + noise Beamforming at BS based on contaminated channel estimation: results in inter-user interference 9

10 Two kinds of approaches Random Access to Pilots Random Access to Pilots and Data Pilot Contention resolution Terminal sends payload when no pilot contention Uplink data is embarked with the pilots Data affected by collision-induced interference Collision in the pilot domain only Collision in the pilot and data domain 10

successful, BS can decode ID of separable users which are

11 Pilot Contamination Suppression Methods based on spatial separation, diversity in path loss, timing offsets If successful, BS can decode ID of separable users which are admitted for DL/UL data transmission UL Pilot transmission 11

12 Pilot Contamination Avoidance Detection of collision at the BS BS sends a message to colliding devices: try again until no collision try again UL Pilot transmission New random access 12

13 Pilot Contamination Avoidance Detection of collision at the device BS sends a precoding pilot 1) Device detects collision 2) Device decides whether to retransmit the pilot sequence Precoded pilot Try again UL Pilot transmission Decision: Retransmit UL Pilot? No collision (with high probability) 13

14 Contaminated channel estimation Pilot Contamination Precoding g = g 1 + g 2 + noise w = g 1+g 2 +noise g 1 +g 2 +noise training y 1 = g 1 H +g 2 H +noise g 1 +g 2 +noise g 1 + n Channel y 2 1 hardening E g 2 i = β i β 1 2 y 2 β 1 +β 2 +σ2 1 = β 1 n β 1 2 Expected Training based Estimate of β 1 + β 2 from y 1 2 : β sum 14 Compare β 1 to β sum Compare β 1 to β sum /2 Collision detection STRONGER user Retransmit the pilot sequence

15 15 Collision resolution

16 Uplink Joint Pilot and Data Transmission For delay-tolerant communications T 1 T 2 T 3 T 4 CW T4 (3) CW T2 (1) p 2 CW T2 (3) CW T2 (L) CW T4 (1) p 2 CW T3 (2) p 2 CW T3 (3) CW T4 (3) p 2 CW T1 (L) CW T4 (L) Time slot τ p τ u One codeword sees an asymptotic number of: channel fades (small and large scale) pilot contamination events 16

17 Uplink Sum Rate Bound MRC at BS τ u τ p τ u E Ka, c K a E = Sum Rate (τ p, p a ) Large scale fading Rate user 0 (K a, c 0, large scale fading) Number of active users Number of contaminators to user 0 g k ~CN(0, β k I) β k 17

18 Uplink Sum Rate Bound- MRC at BS Probability of having Ka active terminals out of K Probability of having c contaminator to a given user conditioned on Ka active users Channel energy K a ~Binomial(K, p a ) c K a ~Binomial(K a 1, 1/τ p ) Optimization wrt τ p and p a 18

19 Heuristic Solution T 1 T 2 T 3 T 4 CW T4 (3) CW T2 (1) p 2 CW T2 (3) CW T2 (L) CW T4 (1) p 2 CW T3 (2) p 2 CW T3 (3) CW T4 (3) p 2 CW T1 (L) CW T4 (L) Time slot τ p τ u τ p o = τ u 3 p a o K = a τ u M Depends on the channel energy variations Sum Rate ~ M τ u 19

20 Average Sum Rate Average sum rate as a function of τ u, K=800 Rate~ M τ u 1 b/s/hz per user M=400 M= b/s/hz per user 20

21 Average Number of Active Users p a K as a function of τ u, K = 800 M=400 M=100 21

22 Conclusions New services and scenarios in 5G: new way to access the pilots and transmit the data Massive MIMO is a fundamental enabler for crowd MBB and mmtc Creation of an efficient standard for wireless networks based on massive MIMO technology will require a complete re-design of the multiple-access layer. 22

Humans and Machines can be Jointly Spatially Multiplexed by Massive MIMO

Humans and Machines can be Jointly Spatially Multiplexed by Massive MIMO Kamil Senel, Member, IEEE, Emil Björnson, Senior Member, IEEE, and Erik G. Larsson, Fellow, IEEE 1 arxiv:1808.09177v1 [cs.it] 28