Joint User Selection and Precoding in Multiuser MIMO Systems via Group LASSO

TL;DR

This paper introduces a group LASSO-based algorithm for joint user selection and precoding in multiuser MIMO systems, achieving higher sum-rates and reduced interference compared to existing methods.

Contribution

It develops a computationally efficient group LASSO approach that improves performance in user selection and precoding for multiuser MIMO systems.

Findings

Higher achievable sum-rate

Lower interference at non-selected users

Enhanced computational efficiency

Abstract

Joint user selection and precoding in multiuser MIMO settings can be interpreted as group sparse recovery in linear models. In this problem, a signal with group sparsity is to be reconstructed from an underdetermined system of equations. This paper utilizes this equivalent interpretation and develops a computationally tractable algorithm based on the method of group LASSO. Compared to the state of the art, the proposed scheme shows performance enhancements in two different respects: higher achievable sum-rate and lower interference at the non-selected user terminals.