RFID Anti-Collision Detection Algorithm Based on Improved Adaptive N-Tree

Abstract

This paper proposes a type of improved adaptive N-tree anti-collision algorithm based on the traditional one for RFID system by combination with maximum likelihood estimation and probe pre-detection. This algorithm inherits some features from Alpha- and tree-based anti-collision algorithms and effectively restrain the star-vation of the two algorithms. It has also filled in the gaps of tag collision with higher probability. The study turns out that the improved adaptive N-tree anti-collision algorithm as proposed can feature adaptive choice of the value N of the tree, length breaks of free timeslots, restraints on defects such as more tag classification and higher collision probability just as what the traditional tree-based algorithm has. N-tree built by level-to-level frame identification eliminates the free timeslots, and improves the tag identification precision for the RFID system. The results from simulation experiment reveal that the algorithm proposed in this paper has lower Error Sampling Reckon (ESR) and Throughput Rate Deviation (TRD), and features large throughput rate (87%), low delay of tag recognition and minimum timeslots, and etc. hence to be better applied in large-scale logistics and other fields where fast information recognition is involved.