A Novel Sensor Node Management Approach Based on Particle Filtering Prediction

In dealing with the moving target tracking in a wireless sensor network (WSN) system, one should not only consider the tracking quality but also the whole system’s energy consumption. However, tracking quality and energy consumption are two incompatible optimization objectives to be achieved simultaneously. Therefore, in the procedure of tracking moving targets, how to select appropriate nodes for carrying out tracking seems to be very important. In this paper, by compromising both tracking accuracy and system energy consumption, a novel node management approach based on particle filtering prediction is proposed. Through the new approach, the moving target’s location at the next moment will be calculated and determined first, then by representing the location as a predefined circular area on the plane, the final tracking nodes can be selected by setting the threshold of the energy consumption and the received signal strength. Computer simulation results show that compared with the traditional selective trigger mechanism based on prediction and the famous probing environment and collaborating adaptive sleeping (PECAS) node scheduling mechanism, the proposed algorithm can not only achieve much better tracking quality, but also activate only a fewer number of nodes for final tracking. Obviously, this advantage will finally reduce the whole WSN system’s power consumption.