Paper Title (use style: paper title)

This paper presents a dynamic particle swarm optimization based search for optimal fusion configuration of sensors in distributed detection network in presence of a nonstationary binary symmetric channel. The wireless channel in sensor networks is a non-stationary random process, which moves the optima of the original problem, otherwise static. The optimal fusion configuration minimizes the probability of error and involves optimal setting of the decision thresholds for the sensors and the optimal fusion rule used at the fusion center. The optimal decision thresholding of a group of sensors in a distributed detection network is a proven intractable problem. Previously a particle swarm optimizer has been used to solve this problem. In this paper a dynamic particle swarm optimizer is used to evolve the optimal decision thresholds of the group of sensors in presence of a non-stationary channel. Performance of the dynamic particle swarm optimizer is compared to the simple particle swarm optimizer in which the particle swarm optimization is restarted after detection of a change. The results demonstrate the effectiveness of the particle swarm in achieving the optima in highly dynamic environments. Dynamic adaptation in the particle swarms towards changes in fitness landscape makes particle swarm optimization an ideal algorithm for multi objective optimization in non-stationary sensor networks.