RFID tag acquisition via compressed sensing: flexibility by random signature assignment

Abstract

Classical Radio Frequency IDentification (RFID) schemes such as Frame Slotted ALOHA (FSA) try to avoid colliding tag responses in the acquisition phase. We propose a scheme that exploits collisions in the acquisition phase, which is enabled by compressed sensing techniques. As an extension to previous work, it allows a more flexible usage of the compressed sensing-based approach for RFID. All activated tags randomly select a signature from a huge signature pool and simultaneously transmit it as a response to a query from the reader. The number of superposed signatures at the reader is significantly smaller than the total signature count, which introduces sparsity to the problem. Utilizing compressed sensing recovery, we determine the set of activated tags, which can then be read out and identified. Our formulation allows to use a computationally efficient approximate message passing algorithm for recovery. The tag identification proves to be quicker and more robust to noise compared to FSA, which we show by simulation.