Toward resilient communication architecture: Online network reconfiguration for UAV failure

Abstract

Given their flexible mobility, rapid deployment capabilities, and cost-effectiveness, unmanned aerial vehicles (UAVs) have recently emerged as a viable solution for providing computational services to distributed devices in the absence of terrestrial infrastructure. However, the vulnerability of the UAV hardware make them susceptible to failures caused by environmental factors and operational stresses, which would directly affect the communication services of these ground users. Therefore, to mitigate potential UAV failures, it is critical to develop a resilient communication architecture that optimizes the configuration of UAVs’ communication networks in an online manner. Consequently, we propose an online network reconfiguration method utilizing the Lyapunov optimization framework to dynamically adjust UAV trajectories and connectivity, thereby minimizing disruptions to communication services in the event of UAV failure. Evaluation results indicate that our proposed method significantly enhances system resilience. In comparison to baseline methods, the proposed method not only maintains high coverage utility and connectivity stability but also minimizes unnecessary UAV movements and operational time costs, thereby demonstrating its superiority.