Autonomous Decision-making Target Tracking Research In Wireless Sensor Networks

Wireless sensor networks (WSNs) is self-organizing networks formed by many tiny low-cost sensors, which is one kind of intelligent network information systems. It is widely used in military and civil fields. As one of the fundamental applications in WSNs, target tracking has attracted great attention from the researchers. Since the hardware resource of sensor nodes are limited, hence how to improve tracking accuracy with low energy consumption becomes the focus of research.Current object localization and tracking Algorithms put too much emphasis on coordinated localization and tracking through communications among nodes, which results in a large energy consumption. On the one hand, this thesis proposes a Locating Edge and its Virtual Node-based Weighted Centroid Target Locating Algorithm, named LEVN-WCTL, which combines the trilateration algorithm with the centroid localization algorithm. The algorithm is based on the polygonal structure to select three nodes nearer to the target, uses the distance from node to the target as the weight and combines the centroid algorithm to estimate the target position. Specifically, we have considered the situation that there are only two locating nodes around the target, and we use the locating edge node and it’s virtual node to estimate the target position. On the other hand, after getting the estimated target position, we propose A Node Sensitivity-based Autonomous Decision-making Target Tracking Algorithm (NS-ADTT for short) on the basis of polygonal tracking structure. In the NS-ADTT Algorithm, the primary endpoint of the locating edge judges whether the target locates in the node adjacent domain or not, and then decides if to participate in the current target tracking without communicating with other nodes.And we compare the NS-ADTT algorithm with tracking algorithm based on the same structure, the simulation results show that the NS-ADTT algorithm can not only guarantee a certain tracking accuracy but also reduce the number of the tracking nodes and the times of the communication between nodes and prolong the network lifetime.