Research On Bath Processing Methods For Large-scale RFID Systems

With the proposal of Internet of Things(IoT),Industrial Internet of things(IIot),the "Made in China 2025" plan,Radio Frequency IDentification(RFID)technology has been one of the core technologies of the networked sensing layer in many wireless network applications,and has been widely concerned by both industrial and academic communities.A number of noval and excellent methods have been proposed in recent years.However,there are still many open issues for existing works.For example,the researching on the large scale RFID system is not mature enough,the agreements and standards of time efficiency have not been unified,the lacks of feasibility considering for the complex functions in various protocols,the shortage of prospective studies for the next generation RFID communication protocol and its application.Focusing on the MAC and application layers of RFID technology,this paper takes the batch processing in the lager-scale RFID systems as the goal,four problems including batch counting,batch authentication,locatilized polling and k nearest neighbor querying are studied in-depth.The detailed content can be briefly described as follows:First,different from existing works which takes the accuracy as the primary goal,we take the time-efficiency rather than accuracy as our final estimation goal,and propose a distributed Bloom Filter based RFID counting method.Based on our experiment observation of the linear relationship between the number of members and the 0/1 numbers in Bloom Filter vector,we establish a two-phases couting model which can greatly reduce the communication overhead between reader and tags,so as to obtain the coarse granularity RFID estimation in constant time.Second,most existing batch authentication methods employ highly complex functions,which is unfeasiable for passive RFID tags.To this,we propose an efficient and lighweight Bloom Filter based batch authentication using Bloom Filter.With two Bloom Filter vectors constructed by genuine tags in database and tags to be verified respectively,we establish the mathematical model to estimate the number of counterfeit tags after a simple combining of these two vectors.We design and implement the batch authentication protocol,which can get the accurate estimation about the number of counterfeit tags for lager-scale RFID systems.Then,because of the unavailability of existing information collection protocols in localized polling,we take the hybrid deployment of general passive tags and passive sensing tags as background,and propose an efficient information collection protocol which consists of two phases: the Tags-Filtering phase and the Ordering-and-Reporting phase.The protocol employs the Bloom Filter twice to significantly reduce the scale of candidate tags in the Tags-Filtering phase.In the Ordering-and-Reporting phase,tags determine their own transmission time-slots according to the allocation vectors which are iteratively broadcasted by the reader.We design and implement the localized polling to achieve efficient and accurate information collection,for all target tags in the current reader's coverage area.Finally,different from most existing query protocols which only can provide the result about the queried tag,we investigate the k Nearest Present Neighbors(kNPN)query problem,which expects to get not only the state of the specified tag,but also the states of its nearest neighbors,especially the k nearest present neighbors,in just one query.To this end,we propose a novel solution in this paper.The pseudo-randomness of hash functions on RFID tags is deliberately amplified and exploited to map each candidate tag to an unique and different slot,so there is no collision at the reader side,and hence is much more efficient than the original methods.