Efficient sensor selection based on spatial correlation in wireless sensor networks

In this paper, we study the selection of active sensors in wireless sensor networks (WSN) subject to a limited energy consumption and a given estimation error. A cost function based on spatial correlation is derived for active sensor selection. Accordingly, a new adaptive algorithm is proposed in which the number of active sensors is adaptively determined and the best topology of active set is selected based on the add one sensor node at a time method. Simulation results show that the distortion of the optimum defined cost function is less than suboptimum one. Also, the performance of the proposed algorithm for sensor selection in computational burden is better than the other algorithm.