Expectation Propagation for Semi-Blind Channel Estimation in Cell-Free Networks

THIS PAPER IS ELIGIBLE FOR THE STUDENT PAPER AWARD. This paper examines the role of Cell-Free (CF) Massive MIMO (MaMIMO) in advancing wireless communication networks, particularly for beyond 5G and 6G networks. Building on the foundational work by Ngo et al., CF MaMIMO, with its distributed architecture, addresses the demands for high data rates, uniform quality of service (QoS), and power efficiency. A central challenge in CF networks is pilot contamination, arising from the absence of traditional cellular boundaries and an excess of user terminals (UTs) relative to pilot sequences. We introduce an Expectation Propagation (EP)-based method for Semi-Blind bilinear estimation in CF MaMIMO networks, providing a low-complexity solution by utilizing the Central Limit Theorem. This method enhances scalability and efficiency compared to existing approaches. Additionally, we propose a shift from distributed to decentralized EP, allowing for local information sharing among Access Points (APs) about user signals.