Design And Implementation Of Data Management System On Heterogeneous Sensor Network

Along with the development and maturation of sensor technology, sensor network have been widely used in many fields, especially in environment monitoring, military applications, medical and health, urban traffic management and smart home. Because of the application range of sensor network becomes more and more wide, application scale becomes more and more large, application requirements become more and more complicate, in order to facilitate the management of sensor network, an efficient sensor network data management system becomes more and more important. There are some problems in the existing sensor network data management systems, for instance, poor portability and scalability, lack of efficient management on heterogeneous sensors.In order to overcome the problems of existing sensor network data management systems, this paper designs and implements a heterogeneous sensor network data management system. This system adopts a distributed and hierarchical architecture, and it divided into four levels from bottom to top:sensor, proxy, control site and client. This architecture is propitious to the deployment and extension of the system.The system uses a protocol independent message communication mechanism, so it will not be affected by underlying routing protocol and can be easily ported to different application environments. The system uses multi-sinks and dynamic data collection method, this can merge the sensor data in temporal and reduce the amount of sensor data transmission, and in the final solve the energy of the network. In data storage and data query, the system uses a distributed method. Sensor data is transmitted to some of the proxies and stored in it, and the commands are divided and send to proxies to execute. The system contains three main functionalities, which are system basic information access, date query and system performance monitoring.The whole system is divided into a number of modules to implement, and different modules are running on different sensors and devices. In order to test the functionality of the system and assess the performance of the system, we deploy a simple heterogeneous sensor network experimental environment and do a large number of demonstration experiments, including the variation of network topology, data queries and performance measurement. This experiments show that the system achieves functionalities and has a good performance.