Research Of Multi-tag Anti-collision Algorithms With Collision Recovery Method

Radio Frequency Identification (RFID) technology is a non-contact automatic identification technology based on radio frequency communication, using radio frequency signal as the carrier of information and energy. With the advantages in identification distance, speed and safety, RFID technology has various applications in the fields of supply chain management, retailing, personal identification, anti-counterfeiting, smart guard, etc. In some emerging applications, high identification speed is required.When two or more tags respond to the reader simultaneously, the reader cannot identify any one of them, and thus multi-tag collision happens. Multi-tag anti-collision algorithms are to solve this issue. In the commonly-used Dynamic-Frame-Slotted-Aloha (DFSM) algorithm, the collision slots are considered useless and the system efficiency cannot exceed 1/e. Anti-Collision algorithm based on collision recovery is a promising method since it can turn collision slots to successful slots.The existing anti-collision algorithms are overviewed, and collision recovery methods are studied. A set of multi-tag anti-collision methods based on collision recovery are proposed, including collision recovery, tag number detection and frame length optimization. Simulation platform based on MATLAB is built to analyze the factors that influence the performance. The hardware architecture with collision recovery is proposed. The system throughput can be promoted above the limitation.According to the simulation results, in two-tag collision slots, the probability to identify both tags is 51%, and the probability to identify either tag is 44%, where the SNR is 30dB. In the scenario of 1000 tags, the system throughput can be promoted from 0.37 to 0.69. With the frame length optimization method in this work, the optimum shrinking factor can be set to 0.5 and the throughput can be further promoted to 0.85. The number of tags in the slots can be detected with the method, and the accuracy can be 90%.The collision recovery module is implemented with SMIC 0.13 μm technology. The area of the module is 0.17mm2. Compared to a digital system without collision recovery, only 10% area is added.