A Low Complexity Detector with Near-ML Performance for Generalized Differential Spatial Modulation

Abstract

Detection of most differential modulation schemes involves the past and present received symbols only. Differential spatial modulation (DSM) is a scheme where an extra degree of freedom for transmitting information is available in the form of transmit antennas. A variant of this scheme called generalized differential scheme for spatial modulation (GD-SM) has a power allocation strategy to improve the error performance as compared to DSM. But the maximum likelihood (ML) detector for this scheme is computationally intensive for higher order modulation. We propose a low complexity detection strategy that makes use of the correlation between the channel coefficients during successive time slots to detect the activated antenna followed by the decoding of the M-ary phase shift keying (MPSK) symbol transmitted through that antenna. The proposed detector achieves a tremendous reduction in complexity close to 83% compared to the ML detector, but with a negligible penalty in error performance.