The growing demand for improved spectral efficiency is one of the main challenges for the upcoming beyond fifth-generation wireless mobile communications networks. While massive multiple-input multiple-output (MIMO) technology has been demonstrating its potential in achieving higher spectral efficiency, the persistent problem of pilot contamination poses a significant hurdle for these systems. To address this issue, the Rate-Splitting Multiple Access (RSMA) framework has emerged as a potential solution. In this paper, we present a novel approach that leverages reinforcement learning (RL) with the Deep Deterministic Policy Gradient (DDPG) algorithm to maximize the sum spectral efficiency (SUM-SE) in a massive MIMO system implementing the RSMA framework with all users sharing a single pilot. The numerical results indicate that the proposed DDPG-based method is a competitive tool for optimizing the SUM-SE in massive MIMO scenarios employing the RSMA framework.
A DDPG-Based Procedure for Mitigating Pilot Contamination in Massive MIMO RSMA Systems
24.06.2024
211315 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch