Collaborative target tracking using distributed Kalman filtering on mobile sensor networks
Reza Olfati‐Saber, Parisa Jalalkamali · 2011
In this paper, we introduce a theoretical frame work for coupled distributed estimation and motion control of mobile sensor networks for collaborative target tracking. We use a Fisher Information theoretic metric for quality of sensed data. The mobile sensing agents seek to improve the information value of their sensed data while maintaining a safe-distance from other neighboring agents (i.e. perform information-driven flocking). We provide a formal stability analysis of continuous Kalman-Consensus filtering (KCF) algorithm on a mobile sensor network with a flocking-based mobility control model. The discrete-time counterpart of this coupled estimation and control algorithm is successfully applied to tracking of two types of targets with stochastic linear and nonlinear dynamics.