A Dynamic HMM-Based Real-Time Location Tracking System Utilizing UHF Passive RFID

Bijoy Dripta Barua Chowdhury, Sara Masoud, Young Jun Son, Chieri Kubota, Russell Tronstad

Research output: Contribution to journalArticlepeer-review

1 Scopus citations

Abstract

In this paper, an indoor real-time location tracking system is proposed using ultra-high frequency passive radio frequency identification technology. A dynamic k nearest neighbor driven hidden Markov model is developed to track subsequent positions of moving objects by combining static localization with dynamic tracking. Statistically validated regression models are developed to relate the tags& x2019; received signal strength indicator and distance, which are iteratively updated over time to accommodate environmental changes. These models are utilized to generate a fingerprint database, which is used by the dynamic k nearest neighbor to estimate the set of potential locations for targeted tagged objects. The probabilistic likelihoods of tags& x2019; potential locations based on the current location and the received signal strength indicator(s) are used to construct the hidden Markov model& x2019;s transition and emission matrixes, respectively. Finally, the performance of the proposed framework is evaluated by a set of experiments given different factors, such as movement pattern and speed. Results reveal that our proposed framework can track objects with an average tracking error as low as 0.36m.

Original languageEnglish (US)
Pages (from-to)41-53
Number of pages13
JournalIEEE Journal of Radio Frequency Identification
Volume6
DOIs
StatePublished - 2022

Keywords

  • Estimation
  • Fingerprint recognition
  • Hidden Markov models
  • Location awareness
  • Passive RFID tags
  • Radiofrequency identification
  • Vehicle dynamics

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Instrumentation

Fingerprint

Dive into the research topics of 'A Dynamic HMM-Based Real-Time Location Tracking System Utilizing UHF Passive RFID'. Together they form a unique fingerprint.

Cite this