Theoretical Analysis Modeling And Adaptive Design Based On MBI Anti-collision Algorithm

As one of the core technologies of Internet of Things,RFID solves the problem of perceptual layer effectively.Besides,it provides effective,accurate and real-time data through linking the material world with the information network.However,the tag collision has a significant impact on the application of large-scale RFID systems.In order to coordinate the transmission of information between tags and readers,it is important to establish a complete and effective anti-collision algorithm.The details of this paper are as follows:Based on the in-depth study of MBI,a model for perfecting the theory of MBI is established from the viewpoint of probability in this paper.Furthermore,an improved Sliding Window Multi-Bit Identification(ISMBI)algorithm is proposed to reduce the communication overhead compared with the conventional MBI algorithm.By means of introducing sliding window and partial bits recovery mechanism,the proposed algorithm can solve the problems of tag response redundancy in regular slot and idle groups in inversion slot simultaneously.In addition,it improves the recognition efficiency of RFID system without increasing any time complexity.In this paper,an Adaptive MBI(AMBI)anti-collision algorithm is proposed to balance the performance of the MBI algorithm.At first,AMBI determines the applicable scenarios of different inversion lengths by analyzing the root cause of the average recognition delay performance of the MBI algorithm.According to the correspondence between the number of tags with the collision bit number and the inversion slot idle group number,Heuristic Function Tag Estimation Algorithm and Idle Group Tag Estimation Algorithm are introduced to estimate the number of tags collectively.Then,according to number of tags,the size of the inversion length is adjusted to achieve the design and implementation of the AMBI algorithm.Obviously,the proposed algorithm can effectively balance MBI algorithm performance and improve system identification efficiency.