Power saving scheduling for multiple target coverage in wireless sensor network

Abstract

In wireless sensor network (WSN), scheduling and power management algorithms have been researched as key techniques to extend network lifetime. In this thesis, we address Multiple Target Coverage Problem (MTCP) in WSN. Conventional sensor scheduling algorithms assume that sensors consume the same transmission energy regardless of how many targets are observed by itself. However, in the MTCP, since the sensor which monitors more number of targets would gather more sensed data, it consumes more transmission energy for sending the collected data to a sink node. Therefore, we assume that the transmitting energy of sensor is proportional to the number of targets covered by sensor itself and its sensing duration. In the MTCP, there also exist overlapped targets that are sensed by adjacent multiple sensors at the same time. Obviously, the overlapped targets are redundancy and result in energy waste due to multiple transmission of the same information. In this thesis, we study on the enhancement of WSN considering the MTCP that has different characteristics unlike the conventional WSN. First, we propose a novel energy efficient scheduling algorithm in the MTCP that considers the transmitting energy in accordance with the number of targets and sensing time. For performance evaluation, we model the optimal sensor scheduling algorithm for MTCP as an Integer Programming that was proved to be NP-complete. To solve the optimal algorithm for MTCP, we propose two heuristic algorithms based on a Linear Programming and a greedy approach which reduce complexity efficiently. Simulation results show that the network lifetime of two heuristic algorithms is close to that of the optimal algorithm for MTCP in case of relatively small number of sensors and targets. Second, we propose an enhanced scheduling algorithm considering overlapped targets in the MTCP. The overlapping sensors, which cover the overlapped target at the same time, gather the same information and send them...