A Data-centric Distributed Detection Scheme in Wireless Sensor Networks

Detection of certain events or targets constitutes an important application of wireless sensor networks (WSN) in which spatially distributed nodes communicate to a decision center. Compared with such decision-centric communication strategies extensively proposed in literature, however, data-centric distributed detection scheme is rather a new design methodology. In this paper, we simulate and determine an efficient way to detect distributed information among sensor nodes. A novel combination of realistic communication channel approximations, Dijkstra's shortest path algorithm, and the optimal ML receiver is introduced to do it efficiently. It is expected and verified through simulation that as the number of nodes in a given area increases, the estimation of the signal at the center node gets better and produces less error