A distributed maximum correntropy Kalman filter

Most distributed Kalman filters are based on the cost function of the well-known minimum mean square estimation criterion, which performs well in the presence of Gaussian noise. When impulsive noise is involved, the performance of distributed Kalman filters may become worse. Recently, a Kalman filter based on the maximum correntropy criterion has been shown to outperform the conventional Kalman filter in the case of impulsive noise. In this paper, we extend the maximum correntropy Kalman filter to a distributed algorithm by introducing a new gain matrix, and analyze the corresponding mean error and mean square error behavior. Simulations are used to demonstrate the effectiveness of the proposed algorithm.