The Research Of The Detection Approaches Of RFID Composite Events Based On Restricted Chronicle

RFID is a non-contact automatic identification technology. In the early stage, itis mainly used in articles identification and tracking. With the application anddevelopment of RFID technology in military, logistics, security and other fields,people use it in the dectection of composite events increasingly, and the technologyis demanded more highly. Traditional chronical detection approaches of compositeevents do not consider the conditions before the dectect registered, which leads tothat the same detections,which are registered in different time, get the differentresults, which does not meet the requirements that whenever the detections areregistered, whenever the detections get the same results.To address this issue, this paper offers an approach of restricted chronicaldetection. The approach uses a certain point of period as a reference, and detectsevents after this reference. When detecting events, the detection using this approachnot only considers the events after the detection was registered, but also consider theevents before registered, which thoroughly resolves the issue that the samedetections,which are registered in different time, get the different resultsTo implement the approach above, this paper offers a detection approachthrough referencing events. This approach refers the results of the same compositeevents which have been registered in the system as the results of the compositeevents registered newly, which reduces the memory consumption, and at the sametime, reduce the computing for events composed.When there are no referencing events in the system, this paper offers a detectionapproach through referencing historical information. According to the historicalinformation, this approach gives the times of events occurring, with which toimplement the detection of composite events based on restricted chronicle.When the detection of composite events are revoked, the correct positioninformation for the detection of composite events will be deleted, which leads to thatthe composite events which are often registered and revoked, will query the historical information all the time. To reduce the times for historical informationquering, this paper organized the historical information which clients care and whichwere ever registered. And this paper offers different modes for different people fordifferent needs.Finally, this paper uses the approaches above to compare with the approachesoffered by the Esper system. A lot of experiments prove that the detectionapproaches of composite events based on restricted chronicle offered by this paperhas a great improvement on the accurate rate of detection and the approaches alsoreduce the use of memory and computing when detecting events.