Real-time data processing based on caching algorithm in Ubiquitous Information Middleware

The infrastructure of ubiquitous computing system is basically formed by WSN (Wireless Sensor Network) and RFID (Radio Frequency Identification) systems. However, applications may have some difficulties in using information obtained from many different types of systems. RFID systems can only provide tag data from readers fixed at a certain location. On the other hand, WSNs deal with seamless data from sensors of which network size is restricted. Therefore, we propose a middleware architecture to integrate WSNs and RFID systems. To integrate two systems, the middleware should support following conditions. The middleware requires common interfaces to communicate with heterogeneous networks. The lack of standard data model means that common data model is needed. In addition, the middleware should be able to provide real-time data for the applications. The proposed middleware uses sensor-agent as a common interface to communicate with different networks. The middleware translates raw data aggregated from sensor-agent into integrated data model. Finally, the middleware has a data caching algorithm based on valid time mechanism. Using the caching algorithm, the middleware can improve the performance of real-time and history data processing. Our experiment shows that the overall workload has been reduced and data can be provided in real-time.