Distributed Antenna Array and RIS-Assisted Planning Framework for Intelligent Coverage Optimization in B5G/6G Cell-Free Massive MIMO.

The transition to Beyond fifth generation of wireless networks (B5G) and sixth generation of wireless networks (6G) exposes the severe interference and coverage limitations of conventional cell-centric architectures. To overcome these bottlenecks, this paper presents a scalable four-layer radio network planning framework that jointly optimizes the deployment of distributed active antenna arrays and passive reconfigurable intelligent surfaces (RISs). The proposed framework integrates a digital twin (DT) loop within an Open-RAN (O-RAN) architecture, employing multi-agent deep reinforcement learning (MADRL) and fractional programming (FP) for real-time joint active and passive beamforming optimization. Extensive Monte Carlo simulations in a dense urban environment demonstrate a 45% increase in spectral efficiency, a 30% reduction in uplink interference, and an 84% reduction in coverage holes compared to legacy 5G networks. Ultimately, these results provide network operators with a cost-effective, standards-compliant blueprint to extend non-line-of-sight (NLOS) coverage by 40% without incurring the prohibitive capital expenditure (CAPEX) of dense active hardware deployments. Furthermore, the proposed architecture demonstrates a competitive 10-15% margin of improvement in spectral efficiency over recent state-of-the-art DRL-based RIS frameworks.

Preparing paper insights from the available abstract and paper details.

Authors

Farre V; Vega-Sánchez J; Cama-Pinto A; Pacheco VG; Orozco Garzón N; Flores-Moyano R

Published

Jan 1, 2026

Journal or venue

Sensors (Basel, Switzerland)

DOI

10.3390/s26154703

Paper type

journal article

Source

Pubmed