Sum Capacity Loss Quantification With Optimal and Sub-Optimal Precoding in Heterogeneous Multiuser Channels

TL;DR

This paper analytically quantifies the expected sum capacity loss between optimal and sub-optimal precoding methods in heterogeneous multiuser channels, considering various transmission scenarios and fading conditions.

Contribution

It introduces a novel affine approximation to predict sum capacity differences between precoders in heterogeneous Ricean fading environments.

Findings

Affine approximation accurately predicts capacity loss across parameters.

Power allocation proportional to user weights maximizes weighted sum capacity.

Numerical results confirm the tightness of the analytical expressions.

Abstract

We analytically approximate the expected sum capacity loss between the optimal downlink precoding technique of dirty paper coding (DPC), and the sub-optimal technique of zero-forcing precoding, for multiuser channels. We also consider the most general case of multi-stream transmission to multiple users, where we evaluate the expected sum capacity loss between DPC and block diagonalization precoding. Unlike previously, assuming heterogeneous Ricean fading, we utilize the well known affine approximation to predict the expected sum capacity difference between both precoder types (optimal and sub-optimal) over a wide range of system and propagation parameters. Furthermore, for single-stream transmission, we consider the problem of weighted sum capacity maximization, where a similar quantification of the sum capacity difference between the two precoder types is presented. In doing so, we disclose that power allocation to different users proportional to their individual weights asymptotically maximizes the weighted sum capacity. Numerical simulations are presented to demonstrate the tightness of the developed expressions relative to their simulated counterparts.