A Non-Alternating Algorithm for Joint BS-RS Precoding Design in Two-Way Relay Systems

TL;DR

This paper presents a non-alternating, low-complexity algorithm for joint precoding design in two-way relay systems that improves fairness without requiring convex solvers or tedious optimization procedures.

Contribution

It introduces a novel method that avoids alternating optimization and convex solvers for joint BS-RS precoding, simplifying the design process.

Findings

Achieves max-min SINR fairness in co-channel users.

Converges to a local optimum without convex solvers.

Validated by simulations showing efficiency.

Abstract

Cooperative relay systems have become an active area of research during recent years since they help cellular networks to enhance data rate and coverage. In this paper we develop a method to jointly optimize precoding matrices for amplify-and-forward relay station and base station. Our objective is to increase max-min SINR fairness within co-channel users in a cell. The main achievement of this work is avoiding any tedious alternating optimization for joint design of RS/BS precoders, in order to save complexity. Moreover, no convex solver is required in this method. RS precoding is done by transforming the underlying non-convex problem into a system of nonlinear equations which is then solved using Levenberg-Marquardt algorithm. This method for RS precoder design is guaranteed to converge to a local optimum. For the BS precoder a low-complexity iterative method is proposed. The efficiency of the joint optimization method is verified by simulations.