Robust Trajectory and Task Allocation in Secure UAV-Assisted MEC System with Cooperative Jamming

The combination of Unmanned aerial vehicles (UAVs) and mobile edge computing (MEC) can efficiently alleviate the computing burden caused by enormous data flow in the context of 6G, especially for the latency-sensitive and computation-intensive tasks in some remote scenarios without available infrastructure. However, the security of information transmitted from ground to air through line-of-sight (LoS) wireless channels remains a major concern due to their broadcast nature. To address this issue, this work studies a secure UAV-assisted MEC system, where a UAV functions as a MEC node to provide computation service for the ground users, while the other UAV functions as a jammer to inject signals to the ground eavesdropper (EV) with an imperfectly known position. The overall energy consumption, encompassing computation, communication, and UAV flight, is minimized under resource limitations. This is achieved through the joint optimization of user transmit power, task allocation for computation, and trajectories for UAVs. Due to the non-convex of the formulated problem, which pertains to the interdependence of various variables in a nonlinear manner, we propose a double-layered robust task allocation and trajectory optimization (RTATO) algorithm. In the first layer, we optimize task and power allocation using the Taylor expansion method. In the second layer, UAV trajectory scheduling is optimized using the slack variable method and successive convex approximation (SCA) method. Simulation results indicate that the algorithm we suggest significantly improves energy consumption in terms of robustness and security compared to benchmark schemes.