Distributed Belief Propagation Detection for User-Centric Cell-Free Massive MIMO

Benefiting from its near-optimal performance and hardware-friendly architecture, Belief Propagation (BP) detection has become an attractive approach for massive multiple-input multiple-output (MIMO) systems. However, its application in the recently emerged Cell-Free massive MIMO (CF-mMIMO) still remains unexplored. In response to this gap, this paper proposes a distributed BP detection algorithm tailored for user-centric CF-mMIMO systems. The proposed detector comprises two modules: 1) Module A for posterior messages updating at each access point (AP), and 2) Module B for computing a priori messages at the central processing unit (CPU). The CPU and APs iteratively exchange the updated messages until reaching the maximum number of iterations. Simulation results demonstrate that the proposed distributed BP detector outperforms the recent art by 7.0 dB SNR gains at 0.1 BER for a CF-mMIMO with 4 distributed APs, each equipped with 8 antennas to serve 8 single-antenna UEs in 16-QAM modulation.