Communication Signal Pre-Processing in Impulsive Noise: A Bandpass Myriad Filtering-Based Method

In impulsive noise modeled as symmetric α-stable (SαS) distribution, it is known that signal processing techniques performed in bandpass can generally achieve better performance than in baseband. In this letter, we propose a bandpass myriad filtering-based pre-processing method for a communication receiver in SαS noise. The adopted myriad filter is designed based on the estimated parameters of SαS noise. The observation data vector, which is used to estimate the transmitted signal, is constructed by choosing specific data from the received bandpass signal, so that two critical requirements for myriad filter can be met. Furthermore, a new method is proposed to estimate the SαS noise parameters, for which the influence of signal is alleviated by an elaborately designed finite impulse response filter. Comparing with the existing methods, numerical simulations validate the superior performance of our proposed bandpass myriad filtering with new noise parameter estimation.