Research On Estimation Of The Number Of Missing Tags For Multi-category RFID Systems

The Radio Frequency Identification(RFID)technology is the integration of a series of electronic technologies,including non-contact,non-line-of-sight,high reliability,ease of deployment and mobility.With the continuous development of Internet of Things(IoTs),the application based on RFID system will be more and more widespread,which is one of the most important auto-identification techniques.Since the system needs to monitor the number of active tags in real time,it is very important to estimate the number of tags in the system quickly and accurately.Based on the existing methods and previous work,this thesis focuses on estimating the number of missing tags of multi-category RFID Systems.At the same time,the noise interference of communication channel is considered.The main contributions of this thesis are as follows:We firstly give the definition of multi-category RFID systems with missing tags.Then,we propose two algorithms named MAC-SZE and MAC-HZE,which are efficiently able to execute missing tags iceberg qeuries for multi-category RFID systems and have high accuracy by reason of directly estimating the number of missing tags.Then we also propose two methods named MAC-SSZE and MAC-SHZE based on the above two estimation algorithms and segmentation ideas.Both methods greatly increase the time efficiency of classification by eliminating unnecessary time slots to reduce the length of frame slots.Simulation results demonstrate that the four methods proposed in this thesis significantly reduce the time cost compared with other methods.We also consider the channel error,which is ubiquitous in RFID systems.The data transmission errors that may occur in the system due to environmental interference.We apply the signal-integrity-check methods such as ECC and CRC in communication principle to the estimation of the number of missing tags.By combined ECC and CRC with the alogrithms proposed in Chapter 3,we extend the alogrithms MAC-SSZE and MAC-SHZE,and the simulation results reveal that the proposed alogrithms can guarantee the predefined accuracy under the environmental interference.Counting the number of present tags and the number of missing tags in large-scale multi-category RFID system is a fundamental problem.Firstly,the ID format of the tag is redesigned in order to satisfy the demand of algorithm,and then the number estimation algorithm of probability geometric distribution is introduced.The algorithm can be applied in RFID system to solve the missing tag estimation problem.Finally,after extensive simulations,we verify the classification accuracy of the algorithm.