BFSearch: Bloom filter based tag searching for large-scale RFID systems

Abstract

Nowadays, RFID technology has been widely applied in inventory management, logistics tracking, identity recognition and so on. In an RFID-assisted large warehouse, customers often require the warehouse to provide the tag searching service to determine which given tags are present in the system. There are two major challenges in tag searching, one is how to exclude the interference caused by unexpected local tags, and the other is how to utilize the known IDs of wanted tags to improve the searching efficiency. In this paper, we propose a Bloom filter based tag searching protocol called BFSearch, which consists of two phases: non-wanted tag deactivation phase and target tag verification phase. In the non-wanted tag deactivation phase, we design a composite filter vector, which can deactivate the local tags that do not belong to the target tag set for eliminating their interference. We further propose an enhanced Bloom filter based tag searching protocol called BFSearch+, which implements a one-to-one correspondence between wanted tags and slots to discover the target tags. It can further improve the time efficiency by up to approximately 50%. Extensive theoretical analysis and simulations illustrate that the proposed BFSearch+ protocol outperforms the state-of-the-art ones.