Predictive power control for dynamic state estimation over wireless sensor networks with relays

We present a predictive power controller for state estimation of a stationary ARMA process over a wireless sensor network (WSN), consisting of sensor nodes, relays, and a single gateway (GW). The state estimate is formed centrally at the GW by using packets received from sensors and relays. The latter perform network coding of sensor measurements. Communication from sensors and relays to the GW is over a fading channel. Packet loss probabilities depend upon the time-varying channel gains and the transmission powers used. To achieve an optimal trade-off between state estimation quality and energy expenditure, in our approach the GW decides upon the in general time-varying transmission powers of sensors and relays. This decision process is carried out on-line and adapts to changing channel conditions by using elements of stochastic model predictive control. Simulations on measured channel data illustrate the performance achieved by the proposed controller.