Real-time Query Processing And Optimization For Basic Events From RFID Data Streams

While RFID (Radio Frequency Identification) is becoming more widely used in various fields, researches on RFID event processing get a high degree of attention. Building platform based on the middleware for the RFID application and building bridge between RFID physical devices and application software become the trend of RFID event processing technology. This processing can be divided into two categories: data-centric processing and event-centric processing. Data-centric RFID event processing technology, because of its performance limitations, is only suitable for an Ad-hoc query or on-line monitoring low-speed events. Event-centric RFID event processing technology is suitable for high-speed event continuous queries, and it has made a lot of innovative technology achievement. In most current RFID complex event monitoring systems, complex event monitors are directly on raw RFID readings. But dealing with a large number of raw data and simultaneously monitoring complex event semantics are very inefficient. Therefore, research on efficient complex RFID event processing techniques has a very important theoretical and practical significance.In order to improve the efficiency of RFID complex event monitoring, we import a data stream management and basic event query processing layer into the RTx-RFID prototype system located between the complex event monitoring and raw data stream. This layer separate the complex event monitoring from the complex raw data stream, making the complex event monitoring only need to monitor the basic event, leaving the data stream management and basic event query processing layer to handle the original data stream. This layer processes according to the subscribed message of the complex event monitoring, generating the basic event, and returns the result to the complex event processing layer for monitoring and processing. This thesis discusses the implementation model of basic event query processing, describes the detailed design of query analyzer and query engine, analyzes a scheduling strategy and studies a operator scheduling algorithm for the basic event query which improves the system's query processing performance.