EUVIMEDEUVIMED entdecken
Crossref

Design and Validation of an RFSoC-Based ISAC System With Real-Time AI-Driven Human Activity Recognition

Sangwoo Shin, Yunhwa Shin, Jaehyun Lee, Jeongsik Choi

IEEE Transactions on Instrumentation and Measurement · 2026 · 75 · 5502016-5502016

DOI: 10.1109/tim.2026.3712911 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Instrumentation and Measurement, 2026. Verfasst von Sangwoo Shin und 3 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 37 Literaturangaben. Gefördert von: National Research Foundation of Korea, Korean Government [Ministry of Science and ICT (MSIT)], Core Research Institute Basic Science Research Program through NRF.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

This article presents the design, calibration, and experimental validation of an integrated sensing and communication (ISAC) measurement system operating at 24 GHz. The system is built on a radio frequency system-on-chip (RFSoC) platform interfaced with millimeter-wave (mmWave) beamformers and is integrated into a software-defined radio (SDR) framework that supports flexible signal processing, real-time data streaming, and diverse experimental configurations. A custom orthogonal frequency-division multiplexing (OFDM) frame structure is designed to embed both data symbols and beam-training sequences within a single transmission, enabling communication and sensing within a single OFDM frame. A systematic hardware delay calibration procedure is also developed to ensure accurate range measurements. Experimental results demonstrate reliable data communication up to 1024-quadrature amplitude modulation (QAM) in an indoor environment. Sensing experiments conducted in an anechoic chamber confirm 2-D target localization with sub-10 cm accuracy using the space-alternating generalized expectation–maximization (SAGE) algorithm. Furthermore, micro-Doppler (MD) signatures extracted from the received signals are processed by a convolutional neural network (CNN) to classify five human activities, achieving over 98 % accuracy in chamber validation and 95.72 % accuracy in indoor room validation. Together, these results demonstrate the potential of the proposed system as a unified and scalable SDR-based platform for end-to-end ISAC prototyping, supporting high-order communication validation, radio frequency (RF) sensing, and artificial intelligence (AI)–driven inference within a single experimental framework.

Förderung

  • National Research Foundation of Korea
  • Korean Government [Ministry of Science and ICT (MSIT)]
  • Core Research Institute Basic Science Research Program through NRF
  • Ministry of Education

Zitieren

APA

Shin, S., Shin, Y., Lee, J., & Choi, J. (2026). Design and Validation of an RFSoC-Based ISAC System With Real-Time AI-Driven Human Activity Recognition. IEEE Transactions on Instrumentation and Measurement, 75, 5502016-5502016. https://doi.org/10.1109/tim.2026.3712911

Vancouver

Shin S, Shin Y, Lee J, Choi J. Design and Validation of an RFSoC-Based ISAC System With Real-Time AI-Driven Human Activity Recognition. IEEE Transactions on Instrumentation and Measurement. 2026;75:5502016-5502016. doi:10.1109/tim.2026.3712911

BibTeX
@article{Shin2026,
  author = {Shin, Sangwoo and Shin, Yunhwa and Lee, Jaehyun and Choi, Jeongsik},
  title = {Design and Validation of an RFSoC-Based ISAC System With Real-Time AI-Driven Human Activity Recognition},
  journal = {IEEE Transactions on Instrumentation and Measurement},
  year = {2026},
  volume = {75},
  pages = {5502016--5502016},
  doi = {10.1109/tim.2026.3712911}
}
Kostenlos starten →