Robust MIMO Receiver Based on Belief Propagation in the Presence of Imperfect Channel and Noise Knowledge

In this paper, the problem of MIMO signal detection based on the belief propagation (BP) algorithm has been addressed in the presence of channel and noise uncertainties at the receiver. We propose a robust detection approach that works on the basis of message passing algorithm on the best tree approximation of the posterior distribution of the received signal. Our approach to make the detector robust is to formulate a worst-case detector design as a min-max optimization problem with imperfect channel and noise knowledge at the receiver.