Real-time Query Processing And Optimising Strategy For RFID Complex Events

As an important technology of the Internet of things, RFID (Radio Frequency Identifier) has been widely applied in various fields. However, RFID data are huge, real-time, implicit and noisy etc. It is difficult to meet the requirements of the application for efficient RFID real-time data processing with the existing traditional database technology and data stream processing technology, so the approach of using event processing technology for RFID data processing is becoming more and more popular in RFID world.In this work, we develop a real-time RFID event processing system based on the characteristics of RFID data, combining with the existing RFID event processing technology. RTx-RFID. RTx-RFID adopts event processing technology and has a three-layer processing structure: Complex Event Processing layer, Basic Event Processing layer and Raw Data Filtering layer. This work mainly focus on Complex Event Processing layer, and designes an SQL-like events query language as user interface, and provides six kinds of algebra operators for event query. Based on query tree model, query sentences are parsed and transformed to Petri nets, which are used in complex event detection. Additionally, Complex Event Processing layer makes much work on comlex event query optimization, including the query algebra rewriting through the adjustment of query tree, sharing the sub query through searching sharing point in the existing query model library. On the other hand, the system proposes a query cost model for every event operation for determining the optimal execution plan.Previous complex event processing mostly focus on the physical optimization of query processing structure, and discuss nothing about logic structure optimization. Some logical optimization strategies are designed in this system.