A team from the Ben Gurion University of the Negev (BGU) in Israel has devised a method for locating the position of drone operators with potentially malicious intentions, flying their drones in protected air space or near airports.

To locate the operators of suspicious drones, the BGU researchers taught a deep neural network, wherein a computer learns to perform specific tasks via training examples, to make predictions about the operator’s location based on the drone’s historical flight path patterns.

During testing, researchers determined that the system could locate the position of a suspicious drone operator 78% of the time following an analysis of simulated flight paths.

Current methods for locating unknown and potentially suspicious drone operators relies on radio frequency techniques and sensors located throughout a flight area that can be triangulated. However, these methods are reportedly affected by nearby Wi-Fi, Bluetooth and internet of things (IoT) signals, possibly interfering with or cloaking drone signals.

The team revealed their findings at the Fourth International Symposium on Cyber Security, Cryptography and Machine Learning (CSCML 2020) on July 3.