Abstract
Radio-frequency identification (RFID) is a RF based identification system, which uses radio waves for communication to exchange data between a reader and an electronic tag attached to an object, for the purpose of identification and tracking. Upon request from reader, tags in reader's accessible RF range will respond, and if the number of responding tags is larger than 2, the reader cannot identify tags (collision state). To avoid the collision, probabilistic approach including ALOHA and deterministic approach including binary tree protocol (BTP) and query tree protocol (QTP) algorithms are presented. However, both of them are essentially collision avoidance algorithms, and require much overhead in retransmission time. In this paper, we present a collision recovery scheme for RFID tag identification system, which does not require additional retransmission for the collision of small number of symbols, which is typical in commercial applications. In our approach, we design a tag collision recovery scheme which is based on (16, 4, 1)-balanced incomplete block design (BIBD) codes. The (16, 4, 1)-BIBD driven symbols generates 20 unique symbols, and we can determine who are colliding RFID tags if the number of collision tags is 2 or 3 (the maximum number of parallelism #3 comes from (16, 4, 1) BIBD parameters). Our experimental results show that we achieved better collision resilience and tag identification performance (even with low SNR environments) than previous approaches..
Original language | English |
---|---|
Pages (from-to) | 1733-1744 |
Number of pages | 12 |
Journal | Information |
Volume | 14 |
Issue number | 5 |
State | Published - May 2011 |
Keywords
- Balanced incomplete block design (BIBD)
- Query-tree algorithm
- Radio frequency identification (RFID)
- Tag collision