Cooperative Algorithms for MIMO Amplify-and-Forward Relay Networks

TL;DR

This paper introduces three cooperative algorithms for MIMO amplify-and-forward relay networks that optimize interference alignment and maximize sum-rate, outperforming existing strategies in various SNR regimes.

Contribution

It proposes novel iterative algorithms for interference alignment and sum-rate maximization in MIMO AF relay networks, with proven convergence and improved performance.

Findings

Higher end-to-end sum-rates achieved compared to existing methods

Algorithms converge to stationary points of optimization problems

Performance varies with SNR and power control strategies

Abstract

Interference alignment is a signaling technique that provides high multiplexing gain in the interference channel. It can be extended to multi-hop interference channels, where relays aid transmission between sources and destinations. In addition to coverage extension and capacity enhancement, relays increase the multiplexing gain in the interference channel. In this paper, three cooperative algorithms are proposed for a multiple-antenna amplify-and-forward (AF) relay interference channel. The algorithms design the transmitters and relays so that interference at the receivers can be aligned and canceled. The first algorithm minimizes the sum power of enhanced noise from the relays and interference at the receivers. The second and third algorithms rely on a connection between mean square error and mutual information to solve the end-to-end sum-rate maximization problem with either equality or inequality power constraints via matrix-weighted sum mean square error minimization. The resulting iterative algorithms converge to stationary points of the corresponding optimization problems. Simulations show that the proposed algorithms achieve higher end-to-end sum-rates and multiplexing gains that existing strategies for AF relays, decode-and-forward relays, and direct transmission. The first algorithm outperforms the other algorithms at high signal-to-noise ratio (SNR) but performs worse than them at low SNR. Thanks to power control, the third algorithm outperforms the second algorithm at the cost of overhead.