Digital forensics for drone data – intelligent clustering using self organising maps

Drones or unmanned aerial vehicles (UAVs) have been rapidly adopted for a range of applications over the past decade. Considering their capabilities of information acquisition and surveillance for intelligence, and increasing access to the common public, have led to a rise in the threat associated with cyber crime associated with drones, in recent times. In order to mitigate the threats and to prevent cyber-crime, digital forensics on drone data is both critical as well as lacking in terms of efficacy studies. In this paper, we define a digital forensic methodology for analyzing drone data, and we propose a self organizing map (SOM)-based method for aiding such analysis. Experiments were conducted on two images obtained from the CFReDS project, namely, ArduPilot DIY Drone and DJI Phantom 4, with the purpose of producing admissible digital forensic evidence for the court of law, as part of a cyber-crime investigation. We also highlight the individual capabilities of several digital forensic tools based on experiments conducted on drone data.