Coordinated multi-point by distributed hierarchical active inference with sensor feedback

Abstract

This study focuses on cooperative beamforming among base stations in wireless communication technology and proposes a new approach based on the Free Energy Principle (FEP). Traditionally, the trade-off between the accuracy of channel information acquisition and overhead has posed a challenge in beamforming. FEP addresses this trade-off by balancing the value from observation and the value from action to select the optimal behavior. This enables adaptive responses to dynamic environmental fluctuations through integrated search, reasoning, and learning. Additionally, by introducing a hierarchical coordination structure, information sharing among base stations and indirect state sharing among agents are achieved, enhancing the efficiency and stability of beamforming. Furthermore, this study utilizes Integrated Sensing and Communication (ISAC) to perform simultaneous communication and real-world sensing. By integrating feedback from terminals and terminal location information for channel state estimation, the overhead is reduced. Simulation results demonstrate that the proposed method effectively improves the Signal to Interference and Noise Ratio (SINR) and energy efficiency.