A Prediction Scheme for Object Tracking in Grid Wireless Sensor Networks

In this study, we use tracking models to predict a moving object´s path in a wireless grid sensor network. The system determines the sleep and awake schedules by using tracking to efficiently conserve energy. At first, the network is divided into several grids by using grid computing. Neighboring grids can communicate with one another and construct an intra-grid power-efficient gathering in sensor information systems (PEGASIS). Furthermore, the sensors may be put into a sleep mode by using the coverage algorithm. Hence, less sensor nodes are in an awake mode and less energy is consumed. After the routing topology construction, the objects will move randomly. We propose a hybrid predictive model, which combines the Grey Theory and Markov chain to predict the moving object´s path. The system determines the sleep and awake schedules for each sensor node. This will decrease the cost which arises from tracking errors and prolong the network´s lifetime.