Bayesian Receiver Design via Bilinear Inference for Cell-Free Massive MIMO with Low-Resolution ADCs

We propose a novel joint channel and data estimation (JCDE) scheme to combat the rate limitation in fronthaul links of cell-free massive MIMO (CF-mMIMO) systems introduced by use analog-to-digital converters (ADCs) at access points (APs), which makes multi-user detection central AP (CAP) challenging. The latter problem is solved here via new JCDE differs from state-of-the-art (SotA) alternatives due two contributions. first design incorporation de-quantization (DQ) step relies only on scalar Gaussian approximation (SGA) assumptions conformity with mild limit theorem (CLT), contrast much harder asymptotic conditions required classic bilinear generalized approximate message passing (BiGAMP) algorithm. second modification belief propagation (BiGaBP), whereby quantized outputs are linearized Bussgang decomposition enabling tractable signal processing. resulting DQ-aided method achieves both low-complexity high-accuracy exploiting spatial degrees freedom (DoF) obtained from, observations CAP compensate for low-resolution distortion by, distributed APs. efficacy proposed over SotA confirmed computer simulations.