A Diffusion Adaptive Network Algorithm for Robust Scalar Field Estimation in Wireless Sensor Networks

In this paper, a robust scalar filed estimation algorithm for wireless sensor networks that contain some nodes with very low quality observations, due to impulsive noise, is proposed. To derive the proposed algorithm, we first recast the robust scalar field estimation as an optimization problem with a maximum correntropy cost function. Then, we develop a distributed solution for it based on the diffusion adaptive network. We evaluate the performance of the proposed algorithm on a scalar field estimation problem and compare it with some similar algorithms, such as diffusion LMS algorithm. Simulation results show the superior performance of the proposed algorithm.