Modulation Diversity in Fading Channels with Quantized Receiver

TL;DR

This paper develops a coding scheme for fading channels with quantized receivers that achieves maximum modulation diversity using low-complexity decoding, even with imperfect channel knowledge.

Contribution

It introduces specific rotations for codewords to ensure modulation diversity in quantized settings and proposes a training and estimation method for practical implementation.

Findings

Achieves maximum modulation diversity with low-complexity decoding.

Proposes a training/estimation scheme that performs well with imperfect channel knowledge.

Error probability similar to perfect channel knowledge scenario.

Abstract

In this paper, we address the design of codes which achieve modulation diversity in block fading single-input single-output (SISO) channels with signal quantization at receiver and low-complexity decoding. With an unquantized receiver, coding based on algebraic rotations is known to achieve modulation coding diversity. On the other hand, with a quantized receiver, algebraic rotations may not guarantee diversity. Through analysis, we propose specific rotations which result in the codewords having equidistant component-wise projections. We show that the proposed coding scheme achieves maximum modulation diversity with a low-complexity minimum distance decoder and perfect channel knowledge. Relaxing the perfect channel knowledge assumption we propose a novel training/estimation and receiver control technique to estimate the channel. We show that our coding/training/estimation scheme and minimum distance decoding achieve an error probability performance similar to that achieved with perfect channel knowledge.