Maximum Likelihood Approach for RFID Tag Set Cardinality Estimation with Detection Errors

Abstract

Estimation schemes of Radio Frequency IDentification (RFID) tag set cardinality are studied in this paper using Maximum Likelihood (ML) approach. We consider the estimation problem under the model of multiple independent reader sessions with detection errors due to unreliable radio communication links and/or collisions. In every reader session, both the detection error probability and the total number of tags are estimated. In particular, after the R-th reader session, the number of tags detected in j ( j=1, 2, ..., R) reader sessions out of R sessions is updated, which we call observed evidence. Then, in order to maximize the likelihood function of the number of tags and the detection error probability given the observed evidences, we propose three different estimation methods depending on how to treat the discrete nature of the tag set cardinality. The performance of the proposed methods is evaluated under different system parameters and compared with that of the conventional method via computer simulations assuming flat Rayleigh fading environments and framed-slotted ALOHA based protocol.