Adaptive Consensus-Based Distributed Kalman Filter for WSNs with Random Link Failures

Abstract

Wireless Sensor Networks have emerged as a very powerful tool for the monitoring and control, over large areas, of diverse phenomena. One of the most appealing properties of these networks is their potentiality to perform complex tasks in a total distributed fashion, without requiring a central entity. In this scenario, where nodes are constrained to use only local information and communicate with one-hop neighbors, iterative consensus algorithms are extensively used due to their simplicity. In this work, we propose the design of a consensus-based distributed Kalman filter for state estimation, in a sensor network whose connections are subject to random failures. As a result of this unreliability, the agreement value of the consensus process is a random variable. Under these conditions, we ensure that the estimator is unbiased, and adaptively compute the gain of the filter by considering the statistical properties of the consensus process. To the best of our knowledge, this is the first time that the design of a consensus-based distributed Kalman filter is addressed by considering the random error introduced by the consensus process. We present some numerical results that confirm the validity of our approach.