低地球轨道卫星边缘计算场景中任务卸载与资源分配联合优化算法

Joint optimization algorithm for task offloading and resource allocation in low earth orbit satellites edge computing scenario

针对低地球轨道(LEO)卫星边缘计算场景中地面用户计算任务的卸载需求,提出联合卸载与资源分配优化(JORAO)算法。考虑到LEO卫星的有限覆盖时间,以最小化所有地面用户的平均服务时延为目标,联合优化卸载策略、LEO卫星的通信和计算资源分配。将任务卸载与资源分配的联合优化问题分解为卸载决策和资源分配子问题,使用交替优化方法,获得原始优化问题的次优解。对于任务卸载决策子问题,将其建模为联盟博弈模型,当博弈达到纳什均衡时,获得最小化平均服务时延的地面用户卸载策略;对于资源分配子问题,使用拉格朗日乘子法获得最优的通信和计算资源分配结果。此外,还证明了所提算法的收敛性和稳定性。仿真结果表明,所提算法具有良好的收敛性,能显著降低地面用户的平均服务时延和提高任务卸载成功率。

Aiming at the offloading requirements of ground users’computing tasks in edge computing scenario of low earth orbit(LEO)satellites,a joint offloading and resource allocation optimization(JORAO)algorithm was proposed.Considering the limited coverage time of LEO satellites,the offloading strategy,the allocation of communication and computing resources of LEO satellites were jointly optimized to minimize the average service delay of all ground users.The joint optimization problem of task offloading and resource allocation was decomposed into offloading decision and resource allocation sub-problems,and an alternating optimization method was used to obtain the suboptimal solution of the original optimization problem.The task offloading decision sub-problem was modeled as a coalition game model,and when the game reached Nash equilibrium,the ground user offloading strategy that minimized the system delay was obtained.For the resource allocation sub-problem,the Lagrange multiplier method was used to obtain the optimal bandwidth and compute resource allocation results.Moreover,the convergence and stability of the proposed algorithm were also demonstrated.The simulation results show that the proposed algorithm has excellent convergence and can significantly reduce the average service delay of ground users,as well as improve the task offloading success rate.