Improving Physical Layer Security for multi-UAV Systems Against Hybrid Wireless Attacks

Multiple unmanned aerial vehicle (UAV) systems are increasingly used in civilian and military applications. However, due to the open and broadcast nature of wireless communications, the control and non-payload communication (CNPC) between UAVs and ground control stations (GCSs) is vulnerable to physical (PHY) and medium access control (MAC) layer wireless attacks, such as eavesdropping, jamming, and spoofing. In this paper, we investigate the secure remote control (RC) issue in multi-UAV systems from a physical layer security (PLS) perspective. We propose a secure RC mechanism based on John Boyd's OODA loop and introduce cooperative jamming (CJ) from UAVs to safeguard control data from GCS. To quantify the uncertainty of wireless attacks from aerial and terrestrial potential attackers, we first introduce the Nakagami-m fading model and a mixed line-of-sight (LoS) / non-line-of-sight (NLoS) probability model to characterize hostile air-to-ground (A2G) channels. We then derive the conditions for successful attacks including passive/active eavesdropping, jamming and spoofing, and formulate the risks of each attack and the overall CJ efficiency based on secrecy outage probability (SOP). To optimize the trade-off between risk costs and CJ power consumption for various situations, we propose three levels of control schemes (centralized, partially centralized, and distributed) applicable at GCS or UAVs, and design three decision-making methods with differing complexities. Through simulations, we evaluate the performance of the proposed mechanism under a general scenario that provides diverse physical conditions, threat levels, and security requirements.