Energy efficient distributed kalman filter for wireless sensor networks

Abstract

Wireless sensor networks (WSNs) have played an important role in collaborative information processing. One of the most fundamental collaborative information processing tasks in WSNs is the distributed estimation for tracking problem. A distributed Kalman filter (DKF) paired with a consensus filter was proposed by Olfati-Saber to address this problem. However, the adjacent nodes need to communicate continuously with each other in the DKF with consensus filter. It was pointed out that transmitting one bit may consume as much power as executing a few thousand instructions. To save the energy of the wireless sensor node, we proposed a DKF with an average algorithm which reduces the sensor computation load and significantly reduces the data exchange between the adjacent nodes compared to the DKF with a consensus filter. A new data exchange protocol is proposed to select the data to exchange between the adjacent nodes. The simulation results show that the proposed DKF with an average consensus filter and DKF with a consensus filter have similar tracking estimation accuracy, whereas the new DKF significantly reduces the level of data transmission.