SNR Estimation in Maximum Likelihood Decoded Spatial Multiplexing

TL;DR

This paper develops SNR estimation methods for maximum likelihood decoded spatial multiplexing in MIMO systems, addressing a key gap in link adaptation by providing practical, low-complexity solutions applicable to modern high-performance communication systems.

Contribution

It introduces novel SNR estimation schemes specifically designed for ML decoded spatial multiplexing, including a low-complexity implementation using the ML decoder itself.

Findings

Proposed SNR estimation methods are effective for ML decoded SM.

The methods are applicable to both horizontal and vertical decoding.

Low overhead implementation with negligible complexity increase.

Abstract

Link adaptation is a crucial part of many modern communications systems, allowing the system to adapt the transmission and reception strategies to changes in channel conditions. One of the fundamental components of the link adaptation mechanism is signal to noise ratio (SNR) estimation, measuring the instantaneous (mostly post processing) SNR at the receiver. That is, the SNR at the decoder input, which is an important metric for the prediction of decoder performance. In linearly decoded MIMO, which is the common MIMO decoding strategy, the post processing SNR is well defined. However, this is not the case in optimal maximum likelihood (ML) decoding applied to spatial multiplexing (SM). This gap is interesting since ML decoded SM is gaining ever growing interest in recent research and practice due to the rapid increase in computation power, and available near optimal low complexity schemes. In this paper we close the gap and provide SNR estimation schemes for ML decoded SM, which are based on various approximations of the "per stream" error probability. The proposed methods are applicable for both horizonal and vertical decoding. Moreover, we propose a very low complexity implementation for the SNR estimation mechanism employing the ML decoder itself with negligible overhead.