An Empirical Study on the Attack and Defense of Unmanned Vehicle

Abstract

One of the main applications of the Cyber-Physical System, the Unmanned vehicle is gradually expanding its use. Unmanned Aerial Vehicle (UAV), among unmanned vehicle, is used not only for cameras, emergency, and military purposes, but its negative effects are increasing also as its use expands. A terrorist outrage using UAVs in Saudi Arabia in the fall of 2019 is a well-known example. Therefore, research on disabling UAV is also becoming important. The UAV neutralization study can be divided into three phases. First, it is the identification of friend or foe stage that distinguishes whether UAVs are friendly or enemy. However, this step can be omitted in No-drone Zones, such as places where people are concentrated, places where major confidential facilities such as nuclear facilities are located, and places of privacy. The second step is to neutralize the UAV's actual mission. At this stage, the UAV is disabled mainly through network attacks such as jamming attacks and packet injection attacks, or through physical attacks such as nets. The third is a post-processing step to lead the UAV to safe area, that is, to prevent the UAV from flying again and to protect the surroundings from it. Previous UAV neutralization studies have focused on disabling UAV without considering the third phase. In this paper, we focused on the third stage, the post-processing stage, so that UAV can be neutralized. Robot Operating System is useful and used widely in UAV system, but there are also vulnerabilities. Therefore, disabling UAVs using this point and defense techniques are discussed in this paper.