Independent component analysis based tag anti-collision algorithm in multi-antenna radio frequency identification

In a radio frequency identification (RFID) system, tag collision is one of key issues for tag identification efficiency. This paper proposes two tag anti-collision algorithms - the ICA-FSA algorithm and the ICA-DFSA algorithm - in a RFID system with dense multi-antenna readers by integrating the ALOHA algorithm and the blind source separation method. In the proposed algorithms, when the number of tags that are transmitting backscattering signals in a specific slot is less than or equal to the number of antennas that a reader connects, the corresponding tags can be identified by using the blind source separation algorithm. Otherwise, these tags cannot be identified, and further select another slot randomly in the same frame to transmit the backscattering signals to the reader according to the ALOHA algorithm until all tags are identified successfully. The simulation results demonstrate that the proposed ICA-FSA and ICA-DFSA algorithms have better performance and less time-consumption than the existing ALOHA-based algorithms, especially when the number of tags increases rapidly.