Multi-Tag Detection Using Multivariate Statistical Analysis for Frequency-Coded Chipless RFID

This paper presents a multivariate statistical scheme to detect multiple chipless radio-frequency identification (RFID) tags. This multivariate statistical technique integrates independent component analysis and time-frequency analysis as a systematic procedure. To demonstrate the proposed technique, we test the read reliability and compare with the direct approach that detects radar cross sections (RCS). The experimental results indicate that the multivariate statistical scheme leads to an improvement of tag recognition. In particular, when the separation between tags is narrow, the multivariate statistical technique enhances the reliability by more than 50%.