Research On Anti-collision And Security Technologies Of RFID System

RFID (Radio Frequency Identification) is known as a non-contact automatic identification technology by exploiting space electromagnetic coupling of radio frequency signals. RFID technology fulfills information input and processing without direct contact, optical visualization and manual intervention. It enjoys many advantages like easy operating, large data storage, good privacy, short response time, high environmental adaptability, far identification distance, strong penetrating ability, multi-object recognition, anti-pollution etc. Currently, it is widely applied in industrial automation, business automation, transportation management, products and documents forgery proofing, theft-proofing and many other fields. With the development of the internet-of-things technology and its applications, RFID, as a key technology, has turned out to be a hot topic in both academia and industry.This dissertation explores the composition and classification of RFID system, the performance of existing anti-collision algorithm for the analysis and simulation, and in-depth research the security and privacy protocol of the RFID system.This paper proposes an improved adaptive multi-tree search (IAMS) anti-collision algorithm by optimizing the prefix sent by reader. IAMS calculates the collision factor and dynamically selects the crossing number of the search tree; it also optimizes the query prefix of quadtree, hence avoiding a lot of idle slots. By means of mathematical analysis, this algorithm computes the required slots which used to identifying tags. The simulation results demonstrate that IAMS has greatly improved the recognition speed and system throughput.Besides, an uninterrupted anti-collision algorithm with ID-based Grouping (UIG) is also proposed. Firstly, it separates tags into different groups resort to the company code and product code. Then, the algorithm generates two initial tag identification codes of each group by the collision information of products'serial number. Finally, it identifies all tags via add or decrease 1 of the initial tags. The analysis of performance and the results of simulation show that the proposed algorithm improves the time complexity and communication complexity, and the throughput is also greatly improved.Based on the Q algorithm and Binary-Tree Algorithm, we propose a novel hybrid algorithm- Enhanced Adaptive Slot-count anti-collision algorithm. It divides the tag into many time slots quickly through improved Q algorithm, and then it identifies tags rapidly with Binary-tree algorithm according to the collision slots stored in the queue. Analysis of performance and the results of simulation reveal that the proposed algorithm has greatly improved the identification efficiency and the throughput. Furthermore, a cyclic shift of the lightweight mutual authentication protocol (CSLMAP) is advanced on the basis of lightweight mutual authentication protocol (LMAP) by rotate left the random number, and then BAN logic is used to prove the safety of the proposed protocol. The results show that the new protocol solves the security and privacy of the LMAP protocol, and improves the efficiency of the authentication protocol, thus reducing the cost of the application of tags.The disseatation also puts forward an improved minimalist mutual-authentication protocol (IM2AP) by protecting the messages transmitted between reader and tag, and then BAN logic is used to prove the safety of the proposed protocol. It is based on minimalist mutual-authentication protocol (M2AP).With the key tag has, IM2AP generates two random numbers which can be shared with reader and then uses the random number to protect the transmitted message on the cyclic shift, thus effectively avoiding the man-in-the-middle attack. By security analysis and performance analysis, this research discloses that this protocol not only maintains the low cost of tags, but also ensures good security and reliability.