Following Targets for Mobile Tracking in Wireless Sensor Networks
Tian Wang,Zhen Peng,Junbin Liang,Sheng Wen,Md Zakirul,Alam Bhuiyan,Yiqiao Cai,Jiannong Cao
DOI: https://doi.org/10.1145/2968450
2016-01-01
ACM Transactions on Sensor Networks
Abstract:Traditional tracking solutions in wireless sensor networks based on fixed sensors have several critical problems. First, due to the mobility of targets, a lot of sensors have to keep being active to track targets in all potential directions, which causes excessive energy consumption. Second, when there are holes in the deployment area, targets may fail to be detected when moving into holes. Third, when targets stay at certain positions for a long time, sensors surrounding them have to suffer heavier work pressure than do others, which leads to a bottleneck for the entire network. To solve these problems, a few mobile sensors are introduced to follow targets directly for tracking because the energy capacity of mobile sensors is less constrained and they can detect targets closely with high tracking quality. Based on a realistic detection model, a solution of scheduling mobile sensors and fixed sensors for target tracking is proposed. Moreover, the movement path of mobile sensors has a provable performance bound compared to the optimal solution. Results of extensive simulations show that mobile sensors can improve tracking quality even if holes exist in the area and can reduce energy consumption of sensors effectively.