Diffusion Normalized Maximum Versoria Criterion Robust to Impulsive Noise

In this brief, we propose a new diffusion normalized maximum Versoria criterion (d-NMVC) algorithm, which is based on maximization of the normalized maximum Versoria criterion (MVC) cost function to enhance the performance of the distributed estimation over networks in the presence of non-Gaussian noise. Convergence of the proposed algorithm, in the mean square sense and evolution behavior, under impulsive noise environment is also analyzed. Simulation results show the robustness of the proposed algorithm under impulsive noise environment against various non-Gaussian noise distributions.