Detection of binary signal with both impulsive and Gaussian interference

In communication environments such as power line channel, impulsive noise greatly deteriorates the performance of detection algorithm based on Gauss optimization. This paper introduces a symmetric alpha stable (SalphaS) distribution to model a statistical model of impulsive noise, and considers that interference in the receiver is a mixture of additional white Gaussian noise (AWGN) and SalphaS noise. Based on this noise model, a nonlinear detector is proposed through the numerical calculation method. Simulations show that the proposed detector is robust, and its performance is slightly worse than the locally optimum (LO) in non-Gaussian environment, however, the key parameters do not need known in advance and its computational complexity is reduced greatly.