Secure Data Transmission Based on Reinforcement Learning and Position Confusion for Internet of UAVs

Ensuring the stability and security of unmanned aerial vehicle (UAV) communication, especially during long-distance missions, is essential for safeguarding against potential attacks. Large-scale UAV communication faces challenges, including eavesdropping threat, data tampering, replay threat, and man-in-the-middle threat. We propose a security information transmission solution based on reinforcement learning and location confusion algorithm (RLPC-SIT) to achieve a secure data transmission between UAVs. First, we leverage the principles of reinforcement learning to identify the most stable transmission routes. Second, we employ location confusion techniques to blur each location of the transmitting UAV with respect to other UAVs. Furthermore, we utilize the concept of message authentication to encrypt the transmitted data, thus making it inaccessible to malicious nodes and preventing forgery. The results of our theoretical analysis and simulation-based experiments indicate that our approach outperforms other security schemes.