Tracking point or diffusing targets using mobile sensor networks under sensing noises

This paper presents a distributed algorithm for a mobile sensor network to track targets with unknown motion. We formulates the target tracking as a multi-objective optimization problem which integrates the tracking quality, the energy saving and the network connectivity. To cope with sensing noises, we use the determinant of the covariance matrix of target estimation as the tracking quality measure and compute its partial derivatives for the optimization process. Virtual nodes are introduced to represent obstacles in the environment. Furthermore this algorithm can be extended to solve the problem of source tracking where sensors can only detect the density of the diffusing substances emitted by the source. Therefore a whole tracking framework has been set up which can be easily extended for applications under complicated situations. Simulations demonstrate the effectiveness of the proposed algorithm in energy conservation and tracking accuracy under different situations.