SDN-Blockchain based Secure Routing for Zero-Trust UAV Networks

In post-disaster scenarios such as earthquakes and wildfires, unmanned aerial vehicles (UAVs) play a crucial role in rapidly restoring communication and assisting rescue operations. However, UAV emergency communication networks face the critical challenges of achieving low-latency, energy-efficient, and secure routing under highly dynamic and threat-prone environments. To address these problems, we design a secure zero-trust architecture (ZTA) by integrating the software-defined networking (SDN) for global routing scheduling and blockchain for tamper-proof security supervision, thereby providing global resource scheduling capabilities for UAV secure routing in emergency communications. Within ZTA, we innovatively define a security degree to quantify the nodes trustworthiness, and provide a unified indicator to evaluate the link safety and guide the routing decisions. Moreover, we design the beam search-proximal policy optimization (BSPPO) routing algorithm to optimize the delay and energy consumption, which supports real-time rerouting upon detecting malicious nodes, ensuring network robustness in hostile environments. In detail, the beam search (BS) is employed to pre-screen candidate nodes and links with high security degree while the proximal policy optimization (PPO) performs per-hop routing decisions. Extensive simulations demonstrate that the proposed BSPPO consistently outperforms baseline algorithms of PPO, BS-Q learning, and BS-actor critic in terms of transmission success rate, delay, and energy consumption, exhibiting superior robustness and adaptability.