Low-Complexity Detection in Large-Dimension MIMO-ISI Channels Using Graphical Models

Abstract

Abstract-In this paper, we deal with low-complexity near-optimal detection/equalization in large-dimension multiple-input multiple-output inter-symbol interference (MIMO-ISI) channels using message passing on graphical models. A key contribution in the paper is the demonstration that near-optimal performance in MIMO-ISI channels with large dimensions can be achieved at low complexities through simple yet effective simplifications/approximations, although the graphical models that represent MIMO-ISI channels are fully/densely connected (loopy graphs). These include 1) use of Markov random field (MRF)-based graphical model with pairwise interaction, in conjunction with message damping, and 2) use of factor graph (FG)-based graphical model with Gaussian approximation of interference (GAI). The per-symbol complexities are ( 2 2 ) and ( ) for the MRF and the FG with GAI approaches, respectively, where and denote the number of channel uses per frame, and number of transmit antennas, respectively. These low-complexities are quite attractive for large dimensions, i.e., for large . From a performance perspective, these algorithms are even more interesting in large-dimensions since they achieve increasingly closer to optimum detection performance for increasing . Also, we show that these message passing algorithms can be used in an iterative manner with local neighborhood search algorithms to improve the reliability/performance of -QAM symbol detection.Index Terms-Factor graphs, graphical models, large dimensions, low-complexity detection, Markov random fields, multiple-input multiple-output inter-symbol interference (MIMO-ISI) channels, pairwise interaction, severe delay spreads.

Channel hardening-exploiting message passing (CHEMP) receiver in large MIMO systems

“…9. Comparison of the BER performance of the proposed CHEMP receiver with those of 1) MMSE detector with MMSE channel estimate, and 2) FG-GAI detector in [11] with MMSE channel estimate, for , 4-QAM.…”

“…We compare the performance of the CHEMP receiver with two other receivers, namely, 1) MMSE detector with MMSE channel estimate, and 2) FG-GAI (factor graph with Gaussian approximation of interference) detector in [11] with MMSE channel estimate. We note that the FG-GAI detector in [11] is also a message passing algorithm which used a Gaussian approximation of interference. But this approximation was done on the original system model in (2), whereas in the proposed MPD algorithm, the Gaussian approximation is done on the matched filtered system model in (7).…”

“…9, we present an uncoded BER performance comparison between 1) proposed CHEMP receiver, 2) MMSE detector with MMSE channel estimate, and 3) FG-GAI detector in [11] with MMSE channel estimate. It can be seen that the performance of the proposed CHEMP receiver is significantly better than those of the MMSE and FG-GAI detectors with MMSE estimate of the channel.…”

“…In the recent years, several low complexity detection algorithms which achieve near-optimal performance in large dimensions using complexities comparable to that of MMSE detection have been proposed [1], [2], [6]- [16]. These algorithms are based on local search (e.g., likelihood ascent search (LAS) algorithm and variants in [1], [2], [6], [7]), meta-heuristics (e.g., reactive tabu search (RTS) and variants in [8], [9]), message passing techniques (e.g., belief propagation (BP) based algorithms in [11], [12]), lattice reduction techniques (e.g., lattice reduction (LR) aided detectors in [13], [14]), and Monte-Carlo sampling techniques (e.g., Markov chain Monte Carlo (MCMC) algorithms in [15]). Issues related to channel estimation and low density parity check codes for large-scale MIMO systems are also being addressed [17], [18].…”

Boosting MMSE Receivers Using AdaBoost

“…It is natural to seek to improve over linear receivers by considering a committee of linear receivers, and given its success story, AdaBoost is a natural candidate to learn the parameters of such a committee. We note that in recent years, ideas from machine learning have been used in communications [18], [19], [20], [21], [22]. In particular, [22] proposes joint channel estimation and multiuser detection based on a "total least squares" approach.…”