A distributed multiple-target identity management algorithm in sensor networks

In this paper, we develop a distributed identity management algorithm for multiple targets in sensor networks. Each sensor is assumed to have the capability of managing identities of multiple targets within its surveillance region and of communicating with its neighboring sensors. We use the algorithm from our companion paper to incorporate local information about the identity of a target when it is available to a local sensor and at the same time reduces the uncertainty of the target's identity as measured by entropy. Identity information fusion is crucial for distributed identity management to compute the global information of the system from information provided by local sensors. We formulate this problem as an optimization problem and present three different cost functions, namely, Shannon information, Chernoff information, and the sum of Kullback-Leibler distances, which represent different performance criteria. Using Bayesian analysis, we derive a data fusion algorithm that needs a prior probability of the given data. Finally, we demonstrate the performance of the distributed identity management algorithm using scenarios from multiple-aircraft tracking in a sensor (radar) network with different fusion criteria.