Interference Alignment Schemes Using Latin Square for Kx3 MIMO X Channel

TL;DR

This paper introduces a Latin square-based interference alignment scheme for Kx3 MIMO X channels, enhancing beamformer selection efficiency and performance with low computational complexity.

Contribution

It proposes a novel Latin square-based IA scheme with a chain structure and a CN-based beamformer selection method for improved performance.

Findings

Larger set of beamformers than conventional schemes

Performance verified through numerical simulations

Low computational complexity of the proposed method

Abstract

In this paper, we study an interference alignment (IA) scheme with finite time extension and beamformer selection method with low computational complexity for X channel. An IA scheme with a chain structure by the Latin square is proposed for Kx3 multiple-input multiple-output (MIMO) X channel. Since the proposed scheme can have a larger set of possible beamformers than the conventional schemes, its performance is improved by the efficient beamformer selection for a given channel. Also, we propose a condition number (CN) based beamformer selection method with low computational complexity and its performance improvement is numerically verified.