Skip to main navigation Skip to search Skip to main content

1D-CNN Enabled Depolarizing Chipless RFID

  • Durham University

Research output: Chapter in Book/Report/Conference proceedingsConference proceedingpeer-review

Abstract

This paper presents the application of a deep learning (DL) regression model to a specialised dataset of depolarised Radar Cross Section (RCS) measurements of a custom 3-bit CRFID tag. A dataset of 12,600 Electromagnetic (EM) RCS signatures were utilized to train and validate a 1-dimensional Convolutional Neural Network (1D CNN) architecture. For the first time, DL implementation approach for reliably detecting identification (ID) information from depolarized signals is presented. In addition, we report the first case of describing the impact of surface permittivity variations, tag deformations, tilt angles, and read ranges, all integrated into model training for enhanced robustness in detecting the encoded tag ID information. It was observed that the developed model was able to generalize well to given data, achieving a low RMSE of 0.040(0.66%) on an unseen test dataset.

Original languageEnglish
Title of host publication2025 IEEE International Conference on RFID Technology and Applications, RFID-TA 2025
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9798331502737
DOIs
Publication statusPublished - 2025
Event2025 IEEE International Conference on RFID Technology and Applications, RFID-TA 2025 - Valence, France
Duration: 28 Oct 202531 Oct 2025

Publication series

Name2025 IEEE International Conference on RFID Technology and Applications, RFID-TA 2025

Conference

Conference2025 IEEE International Conference on RFID Technology and Applications, RFID-TA 2025
Country/TerritoryFrance
CityValence
Period28/10/2531/10/25

Keywords

  • Artificial Intelligence
  • Automation
  • Chipless RFID
  • CNNs
  • Data Acquisition
  • Deep Learning
  • Electromagnetic Systems
  • Machine Learning
  • Radio Frequency
  • RCS
  • RFID
  • Robot

Fingerprint

Dive into the research topics of '1D-CNN Enabled Depolarizing Chipless RFID'. Together they form a unique fingerprint.

Cite this