Distributed Target Tracking in Wireless Sensor Networks With Data Association Uncertainty

This letter considers distributed target tracking in the presence of data association uncertainty. Some known algorithm has been developed where the minimum mean square error approach is utilized to deal with the data association process. However, the computed covariance of global estimate can become non-positive semidefinite when multiple sensors exist, which induces that it can not be performed in some simulations. We develop a novel distributed algorithm where the maximum a posteriori approach is utilized to deal with the data association process. The novel algorithm can ensure that the computed covariance of global estimate is positive semidefinite at all time and it can be performed for all simulations.