Covert attack design on UAV formation in dynamic target tracking tasks

The present paper introduces a covert attack design strategy on Unmanned Aerial Vehicle (UAV) formation, which is mainly applied to the task of tracking dynamic targets. The covert attacks rely on exploiting cyber vulnerabilities, enabling attackers to manipulate GPS readings and inject false information into UAVs through data links. The tracked target in this attack effectively deceives the onboard Global Positioning System (GPS) sensor readings, Laser Distance Sensor (LDS) readings, and manipulates control inputs without altering the network topology of the UAV swarm, thereby inducing increased formation errors between UAVs. Under these conditions, we initially provide the definition and precondition for covert attacks to exist. Subsequently, by transforming the attack problem into a quadratic programming problem with quadratic constraints, we give a design scheme for covert attacks that maximizes the formation error among UAVs. The effectiveness and practicality of the scheme are demonstrated through simulation experiments.