Available

Learning-Based Selective Jamming for Drone Neutralization with Minimal Spectrum Collateral

IoT Security Master Thesis Kirchberg Campus

Overview

Unauthorized drones pose escalating security threats to airports, critical infrastructure, stadiums, and military installations. While counter-drone systems exist, current RF jamming approaches are blunt instruments: they flood the 2.4 GHz and 5.8 GHz bands with wideband interference, disrupting not only the target drone but also WiFi networks, Bluetooth devices, medical equipment, and IoT systems in the vicinity. This collateral damage limits where jamming can be legally and safely deployed.

This project develops intelligent selective jamming that precisely targets drone control links while minimizing impact on neighboring devices. The key insight is that drone communication protocols have distinctive spectral and temporal signatures that differ from standard WiFi/BLE traffic. By learning these signatures, we can craft “surgically” precise interference that disrupts drone-controller communication while leaving other devices largely unaffected.

The approach combines: (1) real-time drone signal classification using deep learning on spectrum observations, (2) protocol-aware jamming waveform design that targets specific drone protocol vulnerabilities (e.g., frequency hopping patterns, ACK mechanisms), and (3) adaptive power control that uses minimum necessary interference.

Ethical Note: This research focuses on defensive applications for authorized counter-drone operators (airports, military, law enforcement) and will be conducted in compliance with spectrum regulations using shielded/anechoic environments or authorized test ranges.

Requirements

Strong programming skills in Python and signal processing.
Background in wireless communications and RF systems.
Understanding of WiFi/BLE protocols and drone communication.
Experience with Software-Defined Radio (GNU Radio, USRP, HackRF, etc.).
Interest in counter-drone systems and spectrum management.

Expected Outcomes

First learning-based selective jamming system for drone neutralization.
Novel protocol-aware jamming waveforms with improved efficiency.
Comprehensive analysis of drone protocol vulnerabilities.
Open-source drone signal dataset and classifier.
High potential for publication at USENIX, CCS, or NDSS.

References

M. Lichtman et al., “A Communications Jamming Taxonomy,” IEEE Security & Privacy, 2016.
P. Nguyen et al., “Matthan: Drone Presence Detection by Identifying Physical Signatures in the Drone’s RF Communication,” ACM MobiSys, 2017.
M. Ezuma et al., “Detection and Classification of UAVs Using RF Fingerprints in the Presence of Wi-Fi and Bluetooth Interference,” IEEE OJCOMS, 2019.
I. Alla et al., “TRIDENT: Tri-modal Real-time Intrusion Detection Engine for New Targets,” Elsevier Computers & Security, 2025.