Learning to Search for MIMO Detection

TL;DR

This paper introduces LISA, a neural network-based iterative search algorithm for MIMO detection that learns optimal decision policies, achieving near-optimal performance and robustness without requiring noise information.

Contribution

The paper presents a novel learning to learn method, LISA, which uses neural networks to optimize decision policies for MIMO detection in various channel conditions.

Findings

LISA achieves near maximum likelihood detection performance.

It outperforms classical and recent deep learning detectors in BER.

LISA is robust to imperfect channel information and generalizes well.

Abstract

This paper proposes a novel learning to learn method, called learning to learn iterative search algorithm (LISA), for signal detection in a multi-input multi-output (MIMO) system. The idea is to regard the signal detection problem as a decision making problem over tree. The goal is to learn the optimal decision policy. In LISA, deep neural networks are used as parameterized policy function. Through training, optimal parameters of the neural networks are learned and thus optimal policy can be approximated. Different neural network based architectures are used for fixed and varying channel models, respectively. LISA provides soft decisions and does not require any information about the additive white Gaussian noise. Simulation results show that LISA 1) obtains near maximum likelihood detection performance in both fixed and varying channel models under QPSK modulation; 2) achieves significantly better bit error rate (BER) performance than classical detectors and recently proposed deep/machine learning based detectors at various modulations and signal to noise (SNR) ratios both under i.i.d and correlated Rayleigh fading channels in the simulation experiments; 3) is robust to MIMO detection problems with imperfect channel state information; and 4) generalizes very well against channel correlation and SNRs.