Free Probability based Capacity Calculation of Multiantenna Gaussian Fading Channels with Cochannel Interference

TL;DR

This paper develops a free probability-based method to accurately calculate the asymptotic capacity of multi-antenna Gaussian fading channels with cochannel interference, considering various impairments and channel configurations.

Contribution

It introduces a generic multi-antenna channel model with four impairments and derives closed-form capacity expressions using free probability theory, extending prior work that ignored interference.

Findings

Derived accurate asymptotic capacity formulas

Validated formulas with numerical simulations

Analyzed impact of cochannel interference on capacity

Abstract

During the last decade, it has been well understood that communication over multiple antennas can increase linearly the multiplexing capacity gain and provide large spectral efficiency improvements. However, the majority of studies in this area were carried out ignoring cochannel interference. Only a small number of investigations have considered cochannel interference, but even therein simple channel models were employed, assuming identically distributed fading coefficients. In this paper, a generic model for a multi-antenna channel is presented incorporating four impairments, namely additive white Gaussian noise, flat fading, path loss and cochannel interference. Both point-to-point and multiple-access MIMO channels are considered, including the case of cooperating Base Station clusters. The asymptotic capacity limit of this channel is calculated based on an asymptotic free probability approach which exploits the additive and multiplicative free convolution in the R- and S-transform domain respectively, as well as properties of the eta and Stieltjes transform. Numerical results are utilized to verify the accuracy of the derived closed-form expressions and evaluate the effect of the cochannel interference.