Research On Offloading Decision-making Mechanism In User Mobile Scenario Of Wi-Fi Network

With the rapid development of 5G technology and various smart terminals,users’ demands for high transmission rates and low latency for data services are increasing.There are also more and more computing-intensive applications,but mobile terminals are constrained by their own computing power and battery capacity,and it is difficult to independently perform high-energy-consuming services locally,such as Virtual Reality(Virtual Reality,VR),Augmented Reality(Augmented Reality,AR),face recognition,etc.Although cloud computing technology can provide users with powerful computing capabilities,long-distance communication between users and cloud servers will cause a large delay,especially for mobile scenarios,which will greatly affect user service experience.Mobile Edge Computing(Mobile Edge Computing,MEC)is an effective solution that provides users with powerful computing power by deploying servers to nodes near users.Compared with MEC based on cellular network,Wi-Fi has the characteristics of large bandwidth,high speed,low cost,easy deployment,and there is no problem of computing charges.Therefore,it is a new idea to use Aps(Access Point,AP)as edge lightweight MEC servers to overcome the computing power and energy consumption limitations of mobile devices.In the scenario of node movement such as the Internet of Vehicles,each node accesses the channel based on competition,and the number of nodes participating in the competition will affect the probability of the node successfully accessing the channel.Moreover,factors such as user moving speed,base station coverage,MEC server computing capability,and the task arrival rate in the MEC server computing cache queue will affect the total delay of offloading tasks.Therefore,this thesis studies the decision-making mechanism of task offloading in the scenario of user mobility in Wi-Fi networks,taking into accounting the delay limitation requirements of offloading services,user moving speed,AP coverage,number of competing access users,MEC server computing capability and The MEC server calculates the impact of parameters such as task arrival rate in the cache queue on system performance.The main research content includes two parts.First,for the multi-user singleAP MEC scenario,a two-level offloading decision-making mechanism executed by the STA(Station,STA)and the AP is proposed.The user satisfaction function is defined according to the STA delay and STA energy consumption,and the optimization goal is to maximize the sum of the satisfaction functions of all users in the system.In order to reduce the complexity of the algorithm,this thesis solves the second level unloading decisionmaking scheme based on the genetic algorithm.The simulation results show that when the number of users is 10,the algorithm can reach convergence within 20 iterations.The performance of the offloading decision algorithm proposed in this thesis is simulated and compared with the performance of three baseline algorithms,which are all local computing,AP offloading decision by probability,and all offloading calculation.The simulation results show that when the number of system users exceeds 20,the algorithm proposed in this thesis can obtain 17%,25% and 36% gains in the task success rate performance that meets the task delay limit,which can significantly improve the sum of the satisfaction function of all users in the system.Secondly,for the multi-AP MEC collaborative user mobility scenario,this thesis considers the impact of user moving speed and AP coverage on the success rate of uplink transmission of computing tasks in the offloading calculation process,and introduces a multi-AP cooperation mechanism to maximize the satisfaction of all users in the system.The sum is the optimization goal,and a two-level offloading decision-making mechanism executed by STA and AP is proposed.Different from the first research content,this research not only decides whether to approve the user’s offloading request,but also decides the AP that feedbacks the result.In this thesis,the genetic algorithm is used to solve the problem,and the simulation results show that when the number of users is 10,the algorithm proposed in this paper can reach stability within 200 iterations.Through the performance simulation analysis and comparison of the algorithm proposed in this thesis and the three defined baseline algorithms.When the number of system users exceeds 20,the offloading decision algorithm proposed in this thesis can reduce the failure rate of tasks exceeding the delay limit by 33%,45%,and 55%,and the system is satisfied Significantly improved.