SAC-based UAV mobile edge computing for energy minimization and secure data transmission

In recent years, the use of UAVs to carry mobile edge computing servers has become an emerging solution to the problem of collaborative ground-to-air communication. However, due to the limited energy consumption of UAVs and the open nature of wireless communication, UAV-based mobile edge computing faces issues such as limited operation time and the potential for insecure connections during data transmission. In this paper, we model a secure data transmission model for UAVs with the goal of minimizing energy consumption. First, we decompose the system energy consumption optimization problem into two sub-problems: ground user energy consumption overhead and UAV energy consumption overhead. We consider the limitations of UAV flight range, computational power, and transmission capability in relation to ground user and UAV, solve the energy consumption optimization problem by deep reinforcement learning, and propose a solution based on the SAC algorithm. The solution applies the idea of maximum entropy to explore the optimal policy and use efficient iterative updates to obtain the optimal policy, enhancing the exploration capability of the algorithm and improving the convergence speed of the training process by retaining all policies with high return values. Secondly, considering the existence of insecure connection nodes in practical application scenarios, a ground user data protection scheme based on Paillier encryption and blockchain is used to achieve secure data transmission. Simulation results show that the proposed method can achieve secure data offloading, effectively reduce the average energy consumption of users, and have good stability and convergence compared with existing methods.