Distributed Sum-Rate Maximization of Cellular Communications with Multiple Reconfigurable Intelligent Surfaces

TL;DR

This paper proposes a distributed optimization framework for maximizing the sum-rate in multi-cell wideband systems using multiple reconfigurable intelligent surfaces modeled as Lorentzian resonators, demonstrating significant performance improvements.

Contribution

It introduces a novel joint design of power allocation and RIS phase profiles considering Lorentzian resonator models, with a distributed solution for multi-cell sum-rate maximization.

Findings

RISs significantly boost system sum-rate.

Distributed optimization effectively manages nonconvex problem complexity.

Simulation results confirm the performance gains of the proposed approach.

Abstract

The technology of Reconfigurable Intelligent Surfaces (RISs) has lately attracted considerable interest from both academia and industry as a low-cost solution for coverage extension and signal propagation control. In this paper, we study the downlink of a multi-cell wideband communication system comprising single-antenna Base Stations (BSs) and their associated single-antenna users, as well as multiple passive RISs. We assume that each BS controls a separate RIS and performs Orthogonal Frequency Division Multiplexing (OFDM) transmissions. Differently from various previous works where the RIS unit elements are considered as frequency-flat phase shifters, we model them as Lorentzian resonators and present a joint design of the BSs' power allocation, as well as the phase profiles of the multiple RISs, targeting the sum-rate maximization of the multi-cell system. We formulate a challenging distributed nonconvex optimization problem, which is solved via successive concave approximation. The distributed implementation of the proposed design is discussed, and the presented simulation results showcase the interplay of the various system parameters on the sum rate, verifying the performance boosting role of RISs.