A Survey on Drone Cybersecurity and the Application of Machine Learning on Threat Emergence

Abstract

Recreational and commercial drones are becoming more common in the United States and globally. As of 2020, recreational drones registered with the Federal Aviation Association (FAA) peaked at 1,195,722 (Federal Aviation Administration, 2021). Threat vectors must be researched, and security concerns mitigated in parallel as drone usage expands. The application of machine learning (ML) is being used to assist in identifying, classifying, and mitigating vulnerabilities and attacks in real time. The emerging threat vector of covert timing channel (CTC) using drones has not been investigated for its feasibility or risk level. This paper summarizes drone vulnerability surveys and seeks to introduce the concept of using ML to assess drone CTC.