Bayesian Method of Slotted Aloha Based Anti-Collision Protocol for RFID Systems

In multiple access systems, knowing the number of tags is crucial for the performance. The problem of multiple access in RFID system has one different assumption from general communication systems in that there is no new user (tag) entering during the conflict resolution interval. This can significantly reduce the uncertainty in number of tags estimation. This paper uses Bayesian method to update the posterior probability distribution of number of users after each slot in Aloha protocol for RFID system. The posterior probability distribution can adapt quickly to concentrate around the true value in few slots. The tag access probability for each slot can also be derived from this posterior distribution to maximize the success probability. The simulation result shows that the system can approach the theoretical limit of the Aloha system with known number of tags.