Intelligent Interference Coordination Mode Selection for Clustered Cell-Free Networks

Yanzan Sun, Tian Bai, Shunqing Zhang, Shugong Xu, Xiaoyun Wang, Shuangfeng Han

Abstract: Clustered cell-free networks provide a scalable architecture where users are dynamically served by selected access points (APs). However, increased AP density intensifies interference management challenges. Current research has predominantly concentrated on Joint Transmission (JT), Coordinated Beamforming (CB), and Non-Cooperation (NC) transmission schemes, which can be viewed as three distinct interference coordination types. Then precoding design and AP-user association optimization under a fixed interference coordination type are widely studied, which we define as interference coordination mode optimization. However, this interference management mechanism limits the overall optimization of interference coordination modes across different interference coordination types, restricting the system’s ability to adapt globally to dynamic network conditions and varying UE requirements. To address this challenge, we propose an Intelligent Interference Coordination Mode Selection (IICMS) scheme based on the multiagent proximal policy optimization (MAPPO) algorithm, specifically designed for clustered cell-free systems operating under backhaul capacity constraints. This scheme effectively facilitates the adaptive selection of interference coordination modes across different coordination types to better respond to fluctuating network conditions and user demands. Subsequently, we adaptively implement corresponding precoding strategies based on the selected interference coordination modes. Simulation results demonstrate that, compared to heuristic adaptive methods and traditional deep learning-based approaches, the proposed IICMS and precoding adaptation significantly reduce user-to-user interference and improve overall system throughput.

External IDs:doi:10.1109/lwc.2025.3581851