Research On Offloading Algorithm In Multi-access Edge Computing Assisted By Vehicles And UAVs

With the rapid development of the fifth-generation mobile communication technology,artificial intelligence applications such as new multimedia services and autonomous driving will become popular choices for mobile users.Different users have different requirements in terms of network delay,throughput,and reliability.Based on service quality and quality of experience,some applications have flexible requirements for network delay and reliability,while some applications have strict requirements.Considering such diverse service quality and experience quality requirements,how to provide users with a better user experience is facing unprecedented challenges in dense user scenarios.A fundamental problem is how to resolve the conflict between computationally intensive applications and resource-constrained mobile devices.It is worth noting that many computationally intensive tasks require a large amount of calculation and require high energy consumption.However,due to physical size constraints,the computing resources and battery life of these lightweight mobile devices are always limited.To solve this problem,mobile edge computing offloading provides a feasible solution.The first part of this paper was to consider task offloading in dense scenarios(such as temporary mass activities,stadium competitions,large shopping malls,etc.).In order to meet the service quality and experience quality requirements of densely distributed users,a multi-access edge computation offloading model with the assistance of vehicles and drones was established.Vehicles,unmanned aerial vehicles,and roadside units acted as edge nodes,different tasks were transmitted to different types of edge nodes for processing as needed.A task offloading scheme based on distributed matching-greedy algorithm was designed to minimize system power consumption,and the impact of the number of users and delay tolerance on the offloading scheme was studied.The simulation results show that the algorithm designed in this paper reduces the power consumption of the system while meeting the delay constraint,and improves the quality of experience of the central user.The second part of this article was to jointly optimize the computing resources and spectrum resources of edge nodes on the basis of the first part.First,the convex optimization algorithm was designed to allocate the optimal computing resources for each user.Secondly,based on delay sensitivity,user satisfaction and resource block quality,a distributed spectrum resource algorithm was introduced.The simulation results show that the designed algorithm achieves the minimum system overhead by effectively allocating transmission resources and computing resources.