Tag collision has a negative impact on the performance of RFID systems. In this letter, we propose an algorithm termed anti-collision protocol based on improved collision detection (ACP-ICD). In this protocol, dual prefixes matching and collision bit detection technique are employed to reduce the number of queries and promptly identify tags. According to the dual prefixes matching method and collision bit detection in the process of collision arbitration, idle slots are eliminated. Moreover, the reader makes full use of collision to improve identification efficiency. Both analytical and simulation results are presented to show that the performance of ACP-ICD outperforms existing anti-collision algorithms.

Fast identification is an urgent demand for modern RFID systems. In this paper, we propose a novel algorithm, access probability adjustment based fine-grained Q-algorithm (APAFQ), to enhance the efficiency of RFID identification with low computation overhead. Specifically, instead of estimation accuracy, the target of most proposed anti-collision algorithms, the APAFQ scheme is driven by updating Q value with two different weights, slot by slot. To achieve higher identification efficiency, the reader adopts fine-grained access probability during the identification process. Moreover, based on the responses from tags, APAFQ adjusts the access probability adaptively. Simulations show the superiority of APAFQ over existing Aloha-based algorithms.