Research And Implementation Of Wireless Sensor Network Analysis And Management Platform

Wireless sensor network is deployed in the monitoring area by a large number of low-cost micro sensor nodes. It is offering many features, accommodating massive nodes, deployed in large area, limited and non-renewable energy, sensing various types of environment information. Analysis of data generated by variety of application and assesses the network status and network performance play an important role in design of WSN application. In this paper, a wireless sensor network analysis and management platform is designed and implemented, based on researching of current visualization system of monitoring WSN.Compared with traditional research work, the WSN platform studied in this thesis is based on―protocol-conversion‖which is able to isolate the lower level hardware feature and upper level software interface. By offering the open platform interface, the platform is able to greatly enhance extensibility and portability of the platform. We also design a data prediction algorithm on the basis of Support Vector Regression (SVR). The algorithm combines the data regularity on the time dimension and data correlation between different nodes, which can solve the problem of unable to assess the status of the monitoring area after some nodes unavailable or failure. Simulation results show that the SVR algorithm can increases the precision of data prediction.Besides, the WSN Analysis and Management platform provides users with multiple-views, such as network topology view, data frame view, real-time data curve and predict data curve. The practical experimental results of the platform indicate that it can present real-time status of WSN topology, various data curves, etc. It can also be deployed as the core platform for Internet of Things (IoT).