Variational Bayesian Inference based Soft-Symbol Decoding for Uplink Massive MIMO Systems with Low Resolution ADCs

Thoota, SS and Murthy, CR (2020) Variational Bayesian Inference based Soft-Symbol Decoding for Uplink Massive MIMO Systems with Low Resolution ADCs. In: Conference Record - Asilomar Conference on Signals, Systems and Computers, 3-6 Nov. 2019, Pacific Grove, CA, USA, pp. 2180-2184.

PDF CON_REC_ASI_CON_SIG_SYS_COM_2180-2184.pdf - Published Version Restricted to Registered users only Download (173kB) | Request a copy

Abstract

In this paper, we present an algorithm to obtain the posterior beliefs of the transmitted bits in an uplink coded massive multiple-input-multiple-output (MIMO) wireless communication system with low resolution analog-to-digital-converters (ADC) at the base station (BS). The nonlinearities introduced by low resolution ADCs necessitates a new multiuser detection approach, for which propose a variational Bayes' inference framework. Further, in coded communications, it is more important to obtain soft symbol estimates for the data bits, rather than hard decisions on the data symbols. We approximate the exact posterior distribution with a fully factorized distribution used in mean field theory in statistical physics, and find its parameters such that the evidence lower bound (ELBO) is maximized. These parameters are used to obtain the posterior beliefs of the transmitted bits in an iterative manner. The resulting algorithm, named the quantized variational Bayesian soft symbol decoding (QVBSSD) is computationally inexpensive and easy to implement. The output posterior beliefs are input to a soft-input-hard-output channel decoder, which gives the decoded bits. We empirically illustrate the bit error rate (BER) of the QVBSSD algorithm, and show its superiority compared to the unquantized MMSE based detection. © 2019 IEEE.

Item Type:	Conference Paper
Publication:	Conference Record - Asilomar Conference on Signals, Systems and Computers
Publisher:	IEEE Computer Society
Additional Information:	cited By 0; Conference of 53rd Asilomar Conference on Circuits, Systems and Computers, ACSSC 2019 ; Conference Date: 3 November 2019 Through 6 November 2019; Conference Code:158954
Keywords:	Analog to digital conversion; Bayesian networks; Bit error rate; Channel coding; Computer circuits; Convolutional codes; Inference engines; Iterative methods; Mean field theory; MIMO systems; Multiuser detection; Statistical Physics, Analog to digital converters; Output channels; Posterior distributions; Soft symbol decoding; Variational bayes; Variational bayesian; Variational Bayesian inferences; Wireless communication system, Decoding
Department/Centre:	Division of Electrical Sciences > Electrical Communication Engineering
Date Deposited:	25 Sep 2020 07:46
Last Modified:	25 Sep 2020 07:46
URI:	http://eprints.iisc.ac.in/id/eprint/65258

Actions (login required)

View Item