A Research On The Localization Technology Of Wireless Sensor Networks

Since many applications of wireless sensor networks depend on the location information of sensor nodes, how to obtain the accurate localization has become the focus of attentions. Self-localization of sensor node is to get the position information of node by using information of some already known nodes and some special mechanisms. Correct self-localization of sensor nodes is the promotion of monitoring events normally. There are still lots of shortcomings in this technology such as great impacted by circumstances, high algorithm complexity and huge energy-costs, unsuitable for mobile sensor networks. we can conclude all of them in one word: unpractical.It is obviously that the technology of wireless sensor networks will go-ahead very quickly and its applications will be wider and wider in the coming future. The research as described in this dissertation is of significance in both theoretical and application areas.This dissertation focuses on the discussion of the practical localization technologies of professional applications based on wireless sensor networks. The major analyses and findings are as follows:1) Design and implementation of A RSSI-based Range-Free Localization Algorithm: By the way of combining RSSI method and Centriod algorithm, the dissertation proposes a new localization algorithm which is less influenced by circumstances. Then discusses the main characteristics of this algorithem, compares it with some other similar algorithms and analyses them. The simulation shows that our algorithm is less influenced by circumstances and 10% better than Centriod-Weighted algorithm.2) Design and implementation of Localization in mobile sensor networks based on Monte Carlo: Under the realistic application background of localization of collier in coal mine, the dissertation gives a new localization algorithm for mobile sensor networks which include: prediction and selection model of contrived moving nodes, definition of filtering rules and analyses with similar localization algorithms. At the end of this part, We validate the practicability of this new algorithm is much better than any other similar algorithms before, it's about 30%～40% better than MCL algorithm and 20%～30% better than CDL algorithm.1) The collier localization and Emergency-aid System of coal mine based on wireless sensor networks: We introduce our research and development on this system simply in this dissertation, including system requirements analyses, introduction of hardware platform, design of system architecture and design of localization sub-system. The realistic application of corresponding algorithms which we talked in the foregoing two parts is explained specially at the end.