Achievable Rates of MIMO Systems with Linear Precoding and Iterative LMMSE Detection

TL;DR

This paper analyzes the achievable data rates in MIMO systems using linear precoding and iterative LMMSE detection, demonstrating potential near-lossless performance with large block lengths and providing theoretical and numerical validation.

Contribution

It introduces a linear precoding method that ensures message Gaussianity and approaches information losslessness in iterative detection for MIMO systems.

Findings

Proposes a linear precoding scheme for MIMO systems.

Shows the scheme is potentially information lossless.

Provides numerical results validating the theoretical analysis.

Abstract

We establish area theorems for iterative detection over coded linear systems (including multiple-input multipleoutput (MIMO) channels, inter-symbol-interference (ISI) channels, and orthogonal frequency-division multiplexing (OFDM) systems). We propose a linear precoding technique that asymptotically ensures the Gaussianness of the messages passed in iterative detection, as the transmission block length tends to infinity. We show that the proposed linear precoding scheme with iterative linear minimum mean-square error (LMMSE) detection is potentially information lossless, under various assumptions on the availability of channel state information at the transmitter (CSIT). Numerical results are provided to verify our analysis.