Multiple Continuous Queries Optimization for Processing RFID Streaming Data

This paper investigates the problem of processing multiple continuous queries over RFID streaming data in RFID middleware. RFID middleware collects and filters RFID sensing data which is continuously generated by RFID readers to process user registered queries. These queries are continuously executed over the incoming tag data to get the matched data. A large number of these queries may be executed simultaneously over a high throughput of incoming data that may deteriorate system performance and cause reporting delay. In order to enhance the performance of middleware, we propose a multiple continuous queries optimization method for RFID middleware, which is based on query execution conditions and filter conditions analysis. Continuous queries having close execution time and large overlapped time intervals are clustered together and an optimal query execution plan is generated for each cluster.