Random sampling algorithm in RFID indoor location system

Abstract

In this paper, low cost radio frequency identification (RFID) indoor location scheme is proposed by deploying RFID tags and implementing a new localization algorithm to make person holding RFID reader know where he is in real time. In this algorithm, the person's state space is represented by maintaining a set of random samples. And a localization method that can represent arbitrary distributions is proposed by using a sampling-based representation. Comparison between proposed algorithm and least square (LS) algorithm in TOA indicated that the positioning errors of the proposed algorithm are lower than LS under the nonline-of sight (NLOS) scenarios. For the case that LS is not available when less than three tags deployed, however, the proposed algorithm can keep track of person.