DoS and DDoS Cyberthreats Detection in Drone Networks

Nowadays, civilian drones are the current trend area and the scientific community continues to come up with new approaches to make this technology more and more sophisticated. In addition, most of the research works conducted by the community mainly focused on a few aspects, namely drone functionality, routing protocols and battery life. However, the study of the computational security posture of drones is ignored and few research works focused on it at a time when cybercriminal actions take advantage of it to compromise the security principles of drone components. Moreover, the geospatial data transiting or stored in drone networks is of a very critical and sensitive nature, as a simple compromise of this data could have adverse impacts that could go as far as endangering human lives, especially in the case of intercepting drones used for blood delivery. Well-known cyber-attacks affecting drone security include Denial of Service (DoS) and Distributed DoS (DDoS) intrusions. DoS and DDoS attacks do not require hacking expertise, and even script kiddies can use tools published on the Internet to implement this type of attack against drone networks. DoS and DDoS attacks can easily result in the loss of the drone and the violation of its security policies. Therefore, protecting drone systems from these intrusions is a vital mission that must be taken seriously to ensure the proper functioning of this technology under safer conditions with a much-reduced risk of falling victim to DoS/DDoS attacks. In this work, we proposed an intrusion detection system capable of detecting known and unknown DoS attacks based on artificial intelligence techniques such as multi-agent systems and machine learning algorithms.