Design And Development Of An RFID Based Complex Event Detection System

As a new technology adopting wireless radio frequency signal to communicate,RFID is a promising electronic identification technology that enanbles a real-timeinformation infrastructure to provide timely,effective object contents for monitoringand tracking applications. It can be used in areas such as supply chian management,pharmaceuticals and quick disbursement and so on.RFID data of objects in RFID system is usually produced in the form of datastream in real time.In some cases, it is even produced in a high speed. For example,RFID technology is used in a real-time monitoring system of emergency treatment. SoRFID data is with the feature of bursting, massive amount. In addition, due tohardware limitations of the device itself, RFID data is uncertainty. Therefore, theproblems to be solved are how to integrate these massive uncertain data to meaningfullogical data and convert to semantic data.In this thesis, we design and implement an RFID based complex event detectionsystem using realistic RFID devices, with the background of object monitoringapplications in hospitals.The system provise a pervasive computing infrastructure forRFID applications. It allows application developers and ordinary users to specifyevents using either a declarative query language or an intuitive visual tool. Based onsimple events, system detects the specified events,forwards them to the registeredapplications and stores them for later use(e.g.,for historical queries). In this scenario,every object is attached to an RFID tag, and the system can monitor the scenario inreal time to recognize the location and activities of objects. Due to using realistic RFID devices, RFID data is uncertainty.and, explorating complex events on theoriginal RFID data is very difficult. So we transfer the original RFID data flow intobasic events flow with probability to manage the RFID data’s uncertainty. And basedon them,we detect the complex events.In short, the system realizes the RFID-based complex event detection system,which exploits a series of high-level data model with semantic information to extractthe complex events.