A blind detector for Rayleigh flat-fading channels with non-Gaussian interference via the particle learning algorithm

A blind particle learning detector (BPLD) is developed for signal detection in Rayleigh flat-fading channels with non-Gaussian interference. The parameters of the fading channel model and the noise model are all unknown. The impulsive noise is modeled as a mixture of Gaussian distributions, which is capable of representing a broad class of non-Gaussian noise. The particle learning algorithm is employed to simultaneously estimate signal and parameters of the fading channel model and the noise model. The delay weight method is used to improve the performance. Simulation results show that the performance of the BPLD proposed can follow closely the performance of the detector with known parameters of the fading channel model and the noise model.