Sequential Parametric Optimization for Rate-Splitting Precoding in Non-Orthogonal Unicast and Multicast Transmissions

TL;DR

This paper introduces FALCON, a sequential parametric optimization method for rate-splitting precoding in non-orthogonal unicast and multicast transmissions, demonstrating near-optimal hybrid precoder performance.

Contribution

The paper proposes FALCON, a novel optimization approach that efficiently solves nonconvex weighted sum-rate maximization for RS precoding without initial feasible point selection.

Findings

FALCON converges to a local optimum reliably.

Hybrid precoders with RS can match fully-digital performance.

Simulations validate the effectiveness of the proposed method.

Abstract

This paper investigates rate-splitting (RS) precoding for non-orthogonal unicast and multicast (NOUM) transmissions using fully-digital and hybrid precoders. We study the nonconvex weighted sum-rate (WSR) maximization problem subject to a multicast requirement. We propose FALCON, an approach based on sequential parametric optimization, to solve the aforementioned problem. We show that FALCON converges to a local optimum without requiring judicious selection of an initial feasible point. Besides, we show through simulations that by leveraging RS, hybrid precoders can attain nearly the same performance as their fully-digital counterparts under certain specific settings.