The Optimal Efficiencyof RFID Tag Identification With Capture Effect

RFID（Radio Frequency Identification）is a non-contact automatic identificationtechnology. It uses radio frequency to communicate, which is fast, efficient, reliable,non-line of sight readable and capable of severer environment. As a key technology ofInternet of Things, RFID has been applied in data collection and goods identification.At present, as the rapid growth of chain number, the function that the RFID readercould read multiple tags simultaneously and efficiently became impending. However,the difference location of tags may leads to the capture effect, which became a hottopic on improving the identification efficiency of RFID system.This thesis studied the RFID anti-collision algorithms concerned with thecapture problem, systematically. The features and innovation points of this thesisinclude:1. A capture-aware tag estimation algorithm named2-D Bayes, whichestimates the tag number and capture probability by the observed slots numberthrough maximizing the posterior probability.2. An estimation method with lowcalculation complexity named CAE, which estimates the tag number by the observedidle slots number and then estimates the capture probability by the estimated tagnumber.3. An algorithm resets the frame length in order to solve the failure of CAEwhile the idle slots number is zero, and it’s named ECAE, which deals with theproblem by extending the frame length through setting the threshold value ofminimum probability and check the idle slots number through bit-slot method.4. Adynamic frame length setting method on the condition of non-equal slots time, whichcopes with ISO18000-6c and EPC C1Gen2standards.5. The dynamic framed slottedALOHAanti-collision protocol with non-equal slots’operation times.