Dynamic Sensor Subset Selection for Centralized Tracking a Time-Varying Stochastic Process

Abstract

Motivated by the Internet-of-things and sensor networks for cyberphysical systems, the problem of dynamic sensor activation for the centralized tracking of an i.i.d. time-varying process is examined. The tradeoff is between energy efficiency, which decreases with the number of active sensors, and fidelity, which increases with the number of active sensors. The problem of minimizing the time-averaged mean-squared error over infinite horizon is examined under the constraint of the mean number of active sensors. The proposed methods artfully combine Gibbs sampling and stochastic approximation for learning, in order to create a high performance, energy efficient tracking mechanisms with active sensor selection. Centralized tracking of i.i.d. process with known distribution as well as an unknown parametric distribution are considered. For an i.i.d. process with known distribution, convergence to the global optimal solution with high probability is proved. The main challenge of the i.i.d. case is that the process has a distribution parameterized by a known or unknown parameter which must be learned; one key theoretical result proves that the proposed algorithm for tracking an i.i.d. process with unknown parametric distribution converges to local optima. Numerical results show the efficacy of the proposed algorithms and also suggest that global optimality is in fact achieved in some cases.