Soft-Input Soft-Output Detection via Expectation Propagation for Massive Spatial Modulation MIMO Systems

In this letter, we propose to utilize the expectation propagation (EP) framework for soft-input soft-output (SISO) signal detection in massive spatial modulation (SM) multiple-input multiple-output (MIMO) systems. By using a multivariate complex Gaussian approximation approach, the symbol beliefs of the transmitted symbol vectors are constructed based on the propagation of the mean vectors and covariance matrixes. The log-likelihood ratios (LLRs) of the antenna index bits and the modulation index bits are computed by partitioning the symbol beliefs into different regions. Simulation results illustrate that the proposed SISO EP detector outperforms the state-of-the-art soft-decision algorithms and strikes a desirable performance-complexity tradeoff.