An EXIT Chart Analysis for Belief-Propagation Based Detection in a Large-Scale MIMO System

Multiple-input multiple-output (MIMO) systems are well known as a method to improve the throughput performance. Recently, a concept of massive (or very large) MIMO using an antenna array with an order of 100 elements has been proposed. In a demand for low-complexity detection method, belief-propagation based algorithm, which is implementable in the order of the second power of the number of elements, has been developed. In the paper, the extrinsic information transfer (EXIT) analysis is applied to clarify the validity of iterative processing capability and evaluate the convergence performance. The results show that the EXIT chart expresses the BER convergence behavior well when the number of antenna elements is 100.