Research On The Detection Of Non-spontaneous Events

As RFID technology becomes more widely recognized,based on its large number of applications:intelligent building management systems,supermarkets or logistics companies supply chain management and so on.RFID system,a large number of basic events for the collection of labels,each label contains information on the ID and the time,place,RFID system is composed by the basic information compound events to represent a variety of semantic complete a variety of upper layer application.Therefore,complex event processing technology as RFID event processing core technology more and more researchers' attention.Which contains a composite event of non-spontaneous events,is important in the composite event detection.How can the efficient detection of composite events,how can the efficient and accurate detection of composite event contains a non-spontaneous events as a difficult RFID event processing.In this paper,the characteristics of the non-spontaneous events in-depth research,in-depth analysis of several typical processing algorithms,a tree-based composite event detection algorithm consists of a non-spontaneous events,the subsequent event-driven composite event detection algorithm.In this algorithm,the first in order to reduce the space complexity of algorithms,only each binocular operator to set up a queue;followed in order to optimize the algorithm's time complexity,change the composite event,in particular,the composite event that contains non-spontaneous events detection mechanism to determine the subsequent events the team into the team to complete the testing;given a set of events under the various operators into the team for the detection of the subsequent event-driven rules.The final experiment showed that our method effectively reduce the algorithm execution time and memory usage.Composite event detection mechanism consists of a non-spontaneous events,also RFID data with uncertainty have been studied and special attention to the uncertainty in the small sample application scenarios.For this paper,RFID composite event of a non-deterministic credibility:Shapiro-Wilk method to determine the distribution of the data,and selected according to the characteristics of the small sample t distribution to calculate the confidence of the data interval,and in order to filter the basic data;data for composite event,in a small sample on the basis of the joint distribution of the mathematical calculations to determine the credibility of the composite event.Finally,comparing the different,past the large sample used in each scene and proprietary algorithms for the application of a small sample of small sample scene to prove the superiority of its proprietary algorithms in the case of the small sample.