Location Algorithm For Wireless Sensor Networks

Wireless sensor networks are a new type of information acquisition and processing technology, and it can be widely used in lots of areas, such as environmental monitoring, military defense and target tracking. The technique of node localization is essential supporting technology in these application areas, and precise positioning is the basis for getting monitoring the location of the event.This paper discussed some related technologies of node localization. Focusing on analysis and research of the DV-Hop, the four shortcomings are as follows:Firstly, because of sensor nodes are deployed randomly, bad nodes appears. Secondly, an average hop distance of anchor nodes take the place of the each jumping distances of unknown nodes and anchor nodes, which results in a greater error. Thirdly, that few of anchor nodes results in greater error and a fewer localization coverage. Finally, trilateration also results in greater computational expense and localization error.To overcome these shortcomings, an improved scheme is proposed. The main work is as follow:First, the measured RSSI value and the mean hop distance are calculated. Then total distance error correction and the average per-hop error correction are obtained using the difference of the measured value and the actual distance. It reduces the impact of positioning accuracy and coverage using the average hop distance error correction.Second, the combination of MIN-MAX method and weighted least squares method take place of trilateration in the process of node localization. First of all, rough estimation position is obtained using MIN-MAX algorithm; secondly, precise position is achieved based on the weighted least squares. Because of reducing floating-point operations, the improved approach does not only reduce the computational expense, but also reduce the average positoning error.It was carried out experiments about the original algorithm and the improved algorithm based on the simulation platform, and it was analyzed and compared based on the different environment parameters of performance affected. The experiment results show that there are more advantages in the aspects of the average positioning error and node location coverage rate.