Research On Multi-uav Computing Offloading Decision Algorithm Based On Mobile Edge Computing

Generally,mobile terminals have limited computing and energy resources,making it difficult to handle tasks with high computing costs and delay-sensitive tasks.The UAV-based Mobile Edge Computing(MEC)system uses the UAV to carry the edge server and transfer the computing task to the edge server to provide on-demand computing services,which provides a solution to the above problems.In addition,drones can establish Line of Sight(Lo S)propagation with ground user equipment to improve task offloading capabilities.Compar ed with the groundbased MEC server,the computing resources carried by a single UAV are limited,the edge computing capability is poor,and the flexibility is low,and it is difficult to cope with the emergency unloading requirements of multi-user terminals.Using multiple drones to form an MEC system can increase the flexibility and capacity of the edge computing system.Under the premise of unified mathematical modeling,this paper optimizes the total mission delay of the multi-UAV MEC system,and carries out the following work:Multiple drones provide computing resources for the ground terminal and form a parallel computing system.Taking minimizing the time required to complete all unloading and computing tasks as the optimization goal,a min-max optimization problem is established to minimize the single drone that takes the longest time to complete the task among all drones.The Rician fading channel model is used to more accurately describe the fading characteristics of the UAV-to-terminal link in three-dimensional space.Model the multi-UAV MEC system based on Time Division Multiple Access(TDMA)transmission scheme and Non-Orthogonal Multiple Access(NOMA)transmission scheme,and introduce corresponding non-convexity for optimization problems Constraints,while jointly optimizing the vertical and horizontal positions of the UAV,as well as the relationship between the UAV and the ground terminal,and unloading the ground terminal in groups.This paper proposes a sub-optimal algorithm based on Successive Convex Approximation(SCA)for solving Mixed Integer Nonlinear Programming(MINLP)problems.The non-convex optimization sub-problem is optimized for solution.Through further performance comparison,the performance advantage of calculating offloading decision based on NOMA transmission scheme is reflected.For UAVs and their ground user groups,a new UAV MEC system calculation and unloading decision-making algorithm is constructed based on uplink NOMA,which uses uplink NOMA to transmit and unload user equipment data in parallel,and optimizes the calculation and unloading decisions in the user group,including users Unloading power allocation,user unloading time slot allocation,etc.,so as to obtain the optimal computing unloading plan,further improving the system’s computing unloading performance,and comparing and analyzing the results of the conventional unloading plan,the results show that the subject plan is compared with the traditional computing unloading Has certain advantages.