Tracking a target with a selected pair of sensors

Maximizing the life of a sensor-network used for tracking the motion of a target depends on selecting a fewer numbers of sensors at a given time. In such applications, it is customary to choose more than two sensors since two-sensor distance measurements entail an ambiguity in two-dimensional position estimation. In this paper, we consider the problem of tracking a mobile target by appropriately selecting two sensors at a given time. The ambiguity in the target position using only two measurements is resolved by a judicious use of the apriori information. Maximum aposteriori (MAP) estimation of the target position in the presence of additive Gaussian noise is developed in the presence of the apriori information. We compute Cramer-Rao lower bound on error variance for estimated location using measurement from n sensors. We discuss how to choose the two sensors from the available set of sensors. Simulation results show that the Newton-Raphson method of finding the zero of a non-linear equation for finding the MAP estimate is a faster alternative to the Steepest descent algorithm commonly used. Simulation results for tracking a mobile target are provided to show the effectiveness of the proposed method.