Fusion Function Placement Algorithm for Distributed Data Fusion Application in Wireless Sensor Networks

Abstract

Distributed data fusion application in Wireless Sensor Networks is very promising, because the change from centralized model to distributed model will average the computation of data fusion algorithm consequently average power consumption in sink node and more importantly placement data fusion functions in networks will decrease data transmission cost and sink node can easily and efficiently re-task different and multiple applications for the same network structure. However, to achieve these advantages data fusion functions need to be placed on optimal node with minimal power consumption and to be dynamically transferred between networks nodes to prevent certain nodes from being exhausted of power. In this paper, a fusion function placement algorithm is proposed for this issue. The simulation results of proposed algorithm and comparison with other algorithm show that our algorithm has better performance including distance to optimal node after initial placement, the delay of fusion function placement and control overhead.