Mobile Sensor Management For Target Tracking

Abstract

In sensor networks, the problem of coverage is a fundamental issue for randomly distributed sensor nodes. In target tracking, it is important to gather a sufficient number of measurements from the sensors to estimate the target trajectory. This paper presents a new approach to improve the tracking accuracy by using mobile sensors with restricted movements. The state of the target and sensors are modeled as a linear Gaussian model and the measurements are assumed non linearly related to the state model and impaired by Gaussian noise. Extended Kalman filtering (EKF) technique is used to estimate the predicted mean square error (MSE) of the estimated target state. We attempt to find the optimal sensor movement and sensor sequence in order to minimize the predicted estimation error subject to satisfying the constraints. Simulation results show that the proposed approach improves the tracking performance