Skip to main navigation Skip to search Skip to main content

Normalized Intrusion Convolution Kernel for Smart Perimeter Detection Using Mach-Zehnder Interferometer Fiber Sensors

  • M. F.F. Pradipta
  • , B. Pamukti
  • , S. Afifah
  • , S. K. Liaw
  • , A. M. Hatta
  • , F. L. Yang
  • National Taiwan University of Science and Technology
  • Academia Sinica - Research Center for Applied Science
  • Institut Teknologi Sepuluh Nopember

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

We present a convolutional neural network-based intrusion detection method called the normalized intrusion convolution kernel, integrated with a Mach-Zehnder Interferometer Optical Fiber sensor. By normalizing vibration signals and applying a convolutional architecture, the system achieves over 99.6% accuracy for real-time smart perimeter intrusion classification.

Original languageEnglish
Title of host publication2025 IEEE Photonics Conference, IPC 2025 - Proceedings
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9798331525590
DOIs
Publication statusPublished - 2025
Externally publishedYes
Event2025 IEEE Photonics Conference, IPC 2025 - Singapore, Singapore
Duration: 9 Nov 202513 Nov 2025

Publication series

Name2025 IEEE Photonics Conference, IPC 2025 - Proceedings

Conference

Conference2025 IEEE Photonics Conference, IPC 2025
Country/TerritorySingapore
CitySingapore
Period9/11/2513/11/25

Keywords

  • Convolutional neural network
  • Distributed optical fiber sensors
  • Intrusion detection system
  • Mach-Zehnder interferometer

Fingerprint

Dive into the research topics of 'Normalized Intrusion Convolution Kernel for Smart Perimeter Detection Using Mach-Zehnder Interferometer Fiber Sensors'. Together they form a unique fingerprint.

Cite this