Research On The Probability Calculation Of RFID Event Based On The Particle Filter

In recent years, radio frequency identification technology(RFID) becomes moreand more popular. It can rapidly obtain information througn the reader and the tag, somore and more application use RFID technology to identify、locate、track and monitorthe object or people and so on.Uncertainty is one of the important features of the RFID system, it exsits duringthe entire life cycle of the RFID application. Usually uncertainty is expressed byprobability value, therefore the expression of the uncertain event is the probabilityevent. How to transform RFID raw data flow into the uncertain event stream(theuncertainty of the basic event is expressed by probability), extracting the compositeevent required, and calculating the probability of the composite event are the focus ofthis study.The work of the paper are as follows:In this paper, the probability of RFID basic events is obtained by the particlefilter, and it proposes a method: firstly,The model of the uncertain basic event is given.The probability table describes the motion direction and the probability accordinglyin the motion model. The particle weight at each location is described by theprobability matrix. At last, the raw data can be transformed to the uncertain basicevent by the probability calculation algorithm of the basic events based on the motionmodel and the weights model.The paper introduces the probability model of the composite event. Thecomposite event is divided into the non-temporal composite event and the temporalcomposite event according to the semantics of the operator. It gives the semanticconversion method, which can caculate the probability of the non-temporal compositeevent consisting with correlated events.The probability of the temporal compositeevent can be obtained according to the theory of conditional probability and theMarkov chain. Finally, the paper gives the detection algorithm of the composite eventbased on the graph, the algorithm mainly introduces the probability calculation of thecomposite event and solves the problem of the semantic ambiguity.