Modified distributed consensus filter for sensor networks

This paper considers distributed consensus filters for sensor networks. In such networks each sensor measures a signal corrupted by noise. We propose an algorithm which significantly improves the accuracy of time-varying signal tracking and attenuates high frequency noise. It is a modification of an existing algorithm. This new algorithm allows us to define the desired bandwidth of the distributed consensus filter by tuning a control parameter. Convergence analysis is provided. Simulation results illustrate the performance of the presented filter.