Coherent MU-MIMO in Block Fading Channels: A Finite Blocklength Analysis

This paper investigates the maximum coding rate of the multiple-user multiple-input-multiple-output (MU-MIMO) uplink in coherent block-fading channels and with finite blocklength. The backoff of the maximum coding rate from the capacity caused by finite blocklength is precisely characterized by a parameter, called the channel dispersion. In particular, we derive exact analytical and approximation results for a large number of the base station (BS) antennas. By analyzing these results, we observe that even when considering the channel dispersion, the maximum coding rate still increases with respect to the number of BS antennas, whilst the SNR for each user can also improve the performance to a ceiling limited by the inter-user interference. Fast channel dynamics (shorter coherence time) and high diversity gain (large number of blocks) are beneficial for the maximum coding rate under finite blocklength and coherent setting. Moreover, to obtain a certain fraction of the capacity with fixed error probability, the minimum required blocklength (delay) can be reduced by increasing the number of BS antennas. We also show that the channel dispersion will converge to a constant while the minimum required blocklength can approach to zero with massive number of the BS antennas. Hence, from a theoretical viewpoint, deploying a large number of the BS antennas is beneficial for low latency communications.

[1]  H. Vincent Poor,et al.  Ultrareliable and Low-Latency Wireless Communication: Tail, Risk, and Scale , 2018, Proceedings of the IEEE.

[2]  Yury Polyanskiy,et al.  Dispersion of the coherent MIMO block-fading channel , 2016, 2016 IEEE International Symposium on Information Theory (ISIT).

[3]  Giuseppe Durisi,et al.  Block-Fading Channels at Finite Blocklength , 2013, ISWCS.

[4]  Mohamed-Slim Alouini,et al.  Capacity of MIMO Rician channels , 2006, IEEE Transactions on Wireless Communications.

[5]  Philippe Ciblat,et al.  Energy-Latency Tradeoff in Ultra-Reliable Low-Latency Communication With Retransmissions , 2018, IEEE Journal on Selected Areas in Communications.

[6]  Vincent Y. F. Tan,et al.  The Dispersion of Nearest-Neighbor Decoding for Additive Non-Gaussian Channels , 2017, IEEE Trans. Inf. Theory.

[7]  Sergio Verdú,et al.  Scalar coherent fading channel: Dispersion analysis , 2011, 2011 IEEE International Symposium on Information Theory Proceedings.

[8]  Emre Telatar,et al.  Capacity of Multi-antenna Gaussian Channels , 1999, Eur. Trans. Telecommun..

[9]  Giuseppe Durisi,et al.  Short-Packet Transmission over a Bidirectional Massive MIMO link , 2019, 2019 53rd Asilomar Conference on Signals, Systems, and Computers.

[10]  Ali Esmaili,et al.  Probability Models in Engineering and Science , 2006, Technometrics.

[11]  Giuseppe Durisi,et al.  Quasi-Static Multiple-Antenna Fading Channels at Finite Blocklength , 2013, IEEE Transactions on Information Theory.

[12]  Erik G. Larsson,et al.  Multipair Full-Duplex Relaying With Massive Arrays and Linear Processing , 2014, IEEE Journal on Selected Areas in Communications.

[13]  H. Vincent Poor,et al.  Channel Coding Rate in the Finite Blocklength Regime , 2010, IEEE Transactions on Information Theory.

[14]  Giuseppe Durisi,et al.  Quasi-static SIMO fading channels at finite blocklength , 2013, 2013 IEEE International Symposium on Information Theory.

[15]  Giuseppe Durisi,et al.  On Single-Antenna Rayleigh Block-Fading Channels at Finite Blocklength , 2017, IEEE Transactions on Information Theory.

[16]  Yury Polyanskiy,et al.  Coherent Multiple-Antenna Block-Fading Channels at Finite Blocklength , 2017, IEEE Transactions on Information Theory.

[17]  Giuseppe Durisi,et al.  Diversity versus channel knowledge at finite block-length , 2012, 2012 IEEE Information Theory Workshop.