Consensus of spatially distributed filters using mobile sensor networks with limited connectivity

The estimation of spatially distributed processes by group of spatially distributed filters utilizing mobile sensors is considered in this work. It is assumed that the sensor network has limited connectivity and a guidance scheme for the mobile sensors is proposed that takes into account the vehicle dynamics. The motion of the mobile sensors is explicitly expressed in terms of the performance of the distributed filters. An added modification into the filter design allows for consensus by penalizing the disagreement of the state estimates in a dynamic manner. Stability bounds are obtained and extensive simulations studies of a representative spatially distributed process are included to provide insights on the effects of moving sensors used in consensus filters.