Research On Highly Efficient Identification Technologies For UHF RFID Systems

Radio Frequency Identification(RFID)is a non-contact automatic identification technology based on inductive or electromagnetic coupling and has become one of the four core technologies supporting the Internet of Things(Io T).With the advantages of mobile,long-distance,and multi-object rapid identification,Ultra-High Frequency(UHF)RFID systems have been widely used in the fields of visual intelligent monitoring and traceability of people and objects.To meet the application requirements of various fields and scenarios,the UHF RFID system needs to collect tag information in time and achieve efficient monitoring of recorded tags.During the process,tag identification technology is the key technology of the UHF RFID system and determines the overall performance.However,traditional tag identification technology cannot efficiently and reliably identify abnormal tags such as new tags,missing tags,and cloned tags among the recorded tag set,and it is also hard to achieve the priority identification and information collection of target tags for a system that contains a large volume of tags.These problems deteriorate the identification efficiency and reliability of the UHF RFID system,which further hinders its wide applications.Although a variety of solutions for abnormal tag identification have been proposed,existing works still suffer from low identification efficiency,poor stability,and high complexity.Therefore,the research on efficient and reliable tag identification technology for UHF RFID systems that meets the requirements of multi-scenario applications has important scientific significance and wide application value.This dissertation focuses on the application scenarios where abnormal tags such as unknown new tags,missing tags,or cloned tags appear and the specific requirements for preferentially identifying some target tags.Based on a systematic in-depth study of tag information extraction,collision bit detection,irrelevant tag filtering,redundant data compression,and other mechanisms,a series of efficient and reliable identification solutions together with the corresponding protocols are proposed.By adopting a combination of theoretical analysis,numerical calculation,and simulation experiments,the performance of the proposed protocols is verified.The main research contents and innovative contributions of this dissertation are as follows:1.To address the problem of long identification time in existing works caused by the inability to efficiently filter out known tags and suffer from several tag collisions,a collision avoidance-based Efficient Unknown(new)Tag Identification(EUTI)protocol is proposed.EUTI builds authentication filters to prevent undesired tags from responding so that all known tags can be quickly deactivated with both expected singleton slots and collision slots.After that,EUTI reduces the tag collision probability by adjusting the frame length and adopts the reservation mechanism to predict the status of subsequent time slots based on the number of collision bits.Accordingly,the new tags can be guided to skip empty slots when responding to the identity information,thus significantly improving the identification efficiency.The simulation results demonstrate that EUTI reduces the known tag deactivation,new tag identification,and total execution time by44.12%,26.47%,and 27.75%,respectively,when compared with the Coding Filtering vector based Unknown tag Identification(CFUI)protocol.2.Aiming at the problems of low time slot utilization and long identification period in existing works caused by ignoring some response information of tags from collision slots,a Sequential String based Missing Tag Identification(SSMTI)protocol is proposed.SSMTI converts singleton and empty slots into collision slots and enables multiple tags in each collision time slot to reply to different strings simultaneously using Manchester encoding.By tracking the collision bits in the aggregated signals,the existence of multiple related tags can be verified together,thus significantly improving time slot utilization and tag identification efficiency.To accelerate the identification process in scenarios of high tag missing rates,an Interactive String based Missing Tag Identification(ISMTI)protocol is further proposed.ISMTI allows some tags to reply with the same string,thereby can identifying more missing tags in collision slots than SSMTI.Moreover,ISFMI can dynamically adjust the tag verification strategy based on the specific missing rate,thus maintaining high identification efficiency when the missing rate changes.Compared with the Collision Resolving based Missing Tag Identification(CR-MTI)protocol,SSMTI and ISMTI respectively reduce the execution time by 45.03% and 68.87%.3.To deal with the problem of low efficiency and poor stability in existing works caused by irrelevant tag interference and tag missing rate changes,an Efficient and Robust missing Key tag Identification(ERKI)protocol is proposed.According to the proportion of key tags,ERKI adopts the improved Bloom filter technology to quickly deactivate irrelevant tags and constructs labeling vectors to optimize the performance of the Bloom filter,thus achieving the rapid separation of key tags and irrelevant tags.Subsequently,ERKI assigns appropriate key tags to respond in the same time slot based on different key tag missing rates.By checking the actual response results,multiple missing key tags can be identified together,thus significantly improving the tag identification efficiency.The simulation results show that the identification speed of ERKI is 2.14 times higher than that of the state-of-the-art Improved VEctor-based missing Key Identification(IVEKI)protocol.4.Aiming at the problem of low identification efficiency in existing works caused by their one-by-one tag verification to find cloned tags,a Cross-Layer Blocked Tag Identification(CLBI)protocol is proposed.CLBI uses the clustering algorithm at the physical layer to estimate the number of tags in each collision slot.By comparing the result with the expected number,it can immediately determine whether clone tags appear in multiple tags,which significantly reduces the number of verifications for tags and thus improves the recognition efficiency.To solve the problem that the performance of CLBI is affected by the proportion of cloned tags,an Adaptive Cross-Layer Blocked Tag Identification(A-CLBI)protocol is further proposed.A-CLBI can estimate the proportion of cloned tags in the system based on the status change of time slots and adaptively adjusts the appropriate tag verification strategy.The simulation results show that A-CLBI reduces the identification time by 49.03% compared with the Enhanced Slotted Broadcastfriendly cloned-tag IDentification(ES-BID)protocol.5.To deal with the problem of high communication overhead and low collection efficiency when adopting existing information collection protocols in multi-category RFID systems,an Arithmetic Coding based Sampling(ACS)protocol is proposed.ACS constructs a sparse vector through the mapping relationship between tags and time slots and uses it to guide the sampled tags from different categories to reply,which not only avoids the interference of irrelevant tags but also reduces the repeated collection of similar information.Moreover,ACS compresses the sparse vector through arithmetic coding,thereby significantly reducing its transmission time.Compared with the Two-Phase Sampling(TPS)protocol,ACS reduces the information collection time by at least 22.70%.This dissertation researches and develops several protocols for the efficient identification of various abnormal tags such as new tags,missing tags,and cloned tags,which support the real-time,high efficiency,and reliability of tag information acquisition.The research results have broadened application prospects in the fields of intelligent monitoring and traceability and are also beneficial to promote the development of UHF RFID technology and industry.