AI Technologies for Drone Operations in Jamming Environments
| Reference number | |
| Coordinator | Wireless P2P Technologies AB |
| Funding from Vinnova | SEK 2 251 560 |
| Project duration | November 2024 - October 2025 |
| Status | Completed |
| Venture | Civil-military synergies |
| Call | Collaborative project for civil-military synergies |
Important results from the project
Objectives were partly met. We completed a feasibility study on an SDR-based cognitive platform and showed PASAD detects intentional jamming on lab data in five classes: clean traffic, AWGN, narrowband pulsed jamming, wideband barrage jamming, and analog VCO jammer. OTA AWGN from Arctic Warrior 2025 corroborated synthetic datasets. Field testing was not reached. Other results: reusable data/ML pipelines, recording/synthesis tools, test harness, and requirements for future threat classification.
Expected long term effects
Results are expected to improve robustness in contested RF environment and strengthen spectrum situational awareness. Methods such as idle/transmission separation and per-band monitoring appear transferable to additional jammer types. The work provides a practical basis to bring the AI technology into operational systems and supports next steps toward field testing and threat classification.
Approach and implementation
The project ran as a feasibility study on an SDR-based cognitive radio platform. We validated PASAD AI technology on lab datasets (clean, AWGN, narrowband pulsed, wideband barrage), then expanded to an analog VCO jammer and a five-class dataset. Two pipelines were used: time-domain idle/transmission separation and frequency-domain per-band FFT monitoring. Over-the-air AWGN recordings from Arctic Warrior 2025 corroborated synthetic data. Full field testing remains for later phases.