Skip to main navigation Skip to search Skip to main content

Radio Frequency Transformer and Reinforcement Learning-Based Adaptive Interference Mitigation for Aviation Communication Systems

  • University of Limerick

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

Abstract

Reliable and interference-free communication is critical for aviation systems operating in complex Radio Frequency (RF) environments. This paper presents an RF-Transformer and Reinforcement Learning (RL)-Based Adaptive Interference Mitigation System to enhance the robustness of aviation communication. A Vision Transformer (ViT), fine-tuned on spectrogram representations of RF signals, is employed for real time interference classification. To mitigate detected interference, a Deep Q-Network (DQN) RF agent dynamically selects optimal countermeasures, including frequency hopping, adaptive filtering, and power adjustment, to enhance SNR and minimize transmission interruptions. The findings validate the framework's capacity to correctly identify interference categories and apply suitable mitigation, supporting reliable communication in aviation settings.

Original languageEnglish
Title of host publicationIrish Signals and Systems Conference
Subtitle of host publicationSignalling our Strength, ISSC 2025
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9798331575939
DOIs
Publication statusPublished - 2025
Event35th Irish Signals and Systems Conference, ISSC 2025 - Letterkenny, Ireland
Duration: 9 Jun 202510 Jun 2025

Publication series

NameIrish Signals and Systems Conference: Signalling our Strength, ISSC 2025

Conference

Conference35th Irish Signals and Systems Conference, ISSC 2025
Country/TerritoryIreland
CityLetterkenny
Period9/06/2510/06/25

Keywords

  • Adaptive Interference Mitigation
  • Aviation Communication
  • Radio Frequency Interference (RFI)
  • Reinforcement Learning (RL)
  • Vision Transformer

Fingerprint

Dive into the research topics of 'Radio Frequency Transformer and Reinforcement Learning-Based Adaptive Interference Mitigation for Aviation Communication Systems'. Together they form a unique fingerprint.

Cite this