Tracking and Prediciting Movement Patterns of a Moving Object in Wiresless Sensor Network

Abstract

One of the recent applications in wireless sensor network is to detect and predict the movement of a moving object in Wireless Sensor Network (WSN). To trace the moving object we propose a FaceTrack framework in which the region is divided into different shaped polygons called Faces. After tracing the path of the moving object; the future path is predicted using Apriori Algorithm and Pattern Recognition (PR) Algorithm. In Apriori Algorithm; data mining is performed on the past movement information of the moving object, from which the association rules are excavated, which are used to predict the next location of the obj ect. In PR algorithm we use the current movement pattern of the obj ect, and then we should be able to predict the next location of a moving object in sensor networks and to activate the least sensor nodes. In PR algorithm our policy is to take the sensor node with highest frequency as the first predicted node based on the movement patterns discovered by motion pattern generation. If the prediction fails, we extend the region consisting of the cluster containing the predicted sensor node in the higher level and wake up all sensor nodes within this region until the moving object is found. In the worst case, we activate all sensor nodes in the network.