Robust multiuser detection in flat fading non-Gaussian channels

This paper deals with the problem of multiuser detection (MUD) in direct-sequence code-division multiple-access (DS/CDMA) fading channels with impulsive noise. This issue is of importance because in many practical situations the ambient noise is impulsive due to various natural and man-made impulsive sources. In this paper a novel H^∞ estimation based robust multiple model detector (RMMD) is proposed. We have considered here binary phase-shift keyed (BPSK) signals transmitted simultaneously by multiple users via a CDMA flat-fading channel and embedded in impulsive noise. Simulation results show that, in impulsive noise and low signal-to-noise ratio (SNR), the bit-error-rate (BER) performance gain afforded by the proposed multiuser detector can be substantial when compared to a H² based multiple model detector (IMM-SIC) and the iterative M-estimator-based decorrelating detector, recently proposed in the literature.