A Novel Parameters Estimation Scheme for Multivariate Symmetric Alpha Stable Noise in High-Order Modulation System

The impulsive interferences are not negligible in many communication channels, such as power line channels. During the initialization of Turbo iterative decode algorithm, e.g., message passing algorithm for Low-Density Parity-Check code's decoder, it is necessary to estimate parameters of channel to retrieve the prior information from the received signals. Performances of the detector/estimator based on Gaussian optimization are deteriorated sharply due to impulsive noise. It does make sense to restudy how to get the prior information in impulsive channel. We introduce symmetric alpha-stable (SocS) noise into a statistical model of impulsive noise, and consider that interference in the receiver is a mixture of multivariate SocS noise. A novel parameters estimation scheme for impulsive channel in high-order modulation scenarios is proposed, which is based on equivalent characteristic function. It proves to be effective and low complexity for both MPSK and MQAM, as we can see in the simulations.