초록 close

U-Environment Monitoring System using USN effectuates real-time remote monitoring by way of pollution measuring instruments and sensors in anytime, anywhere. It allows preventing the spread of contamination by early examination of various environmental conditions. But it's difficult to handle vast amounts of data obtained from numerous sensors in a single computer system; it demands more storage resources. By default, in a single sensor pH, DO, temperature, etc. When measuring, sensing the amount of data that occurred on average per minute is approximately 1,000 bytes. In other words, in real time to process large amounts of data, problems arise. Thus, studies on greed computing are necessary in order to connect resources such as geographically dispersed computers, networks, servers and to use them as a single resource. Data grid is a technology that permits efficient sharing of data concentrated in one place and accessing large amounts of data and DB in real time across multiple locations. This paper develops middleware based on grid data and applies it to u-Environment Monitoring System. The middleware proposed in this paper enables abstraction of heterogeneous sensors and improvement of processing speed by distributing vast amounts of data. In case of the network overhead or server failure, it maintains continuity of operations by distributing the data to another server associated with a grid network.