Research On Optimization Technologies Of Task Scheduling In Mobile Edge Computing

Mobile edge computing(MEC)has now become a new service paradigm by its low latency and high performance.Under this paradigm,mobile devices(e.g.,smartphones)can offload computation-intensive workloads to edge servers nearby through WiFi or cellular networks.However,with the development of MEC,how to more safely and efficiently unload user tasks and schedule tasks to a suitable MEC server has become a popular research point of MEC nowadays.Therefore,this paper focuses on the task scheduling problem in mobile edge computing.Starting from task scheduling security and efficiency,comprehensively considering system performance and power consumption,this paper design and implement a performance and power consumption trade-offs task scheduling with trust-guaranteed algorithm and an optimization algorithm for task offloading and scheduling trade-offs.Fristly,based on the security problem of task scheduling on the MEC platform,and comprehensively considering the conflicting trade-off between system performance and energy consumption,this paper uses Lyapunov optimization to propose a performance and power consumption trade-offs task scheduling with trust-guaranteed algorithm DOSSTM.It designs a light-weighted trustworthiness management strategy,which can resist fake QoS attacks and regulate the server providers’ behavior.DOSSTM uses a distributed approach to decompose the problem into three sub-problems and proposes three online strategy:(1)a“Join the Best Queue” request routing strategy for balancing the system profit and queue backlog without a priori knowledge of requests arrivals;(2)a “Real-time Dynamic Adjustment” energy management strategy for controlling a running state of a virtual machine to reduce power consumption;(3)a light-weighted trustworthiness management strategy for calculating the trustworthiness value to regulate the candidates’ behavior.Mathematical analyses and simulations have demonstrated that the performance and power consumption trade-offs and trusted QoS management can be achieved simultaneously by DOSSTM.Secondly,based on the efficiency of the MEC platform’s task offloading and task scheduling trade-offs,and comprehensively considering the conflicting trade-offs between system utility and energy consumption,this paper uses Lyapunov’s optimization to propose an online edge server scheduling algorithm for optimal system utility that guarantees the load balance of smart mobile devices and edge servers,called DOMES.Specifically,DOMES has established a double-queue model to describe the workload of mobile users and edge servers.The algorithm decomposes the problem into three sub-problems in a distributed manner and proposes three online strategy:(1)a “Optimal Bandwidth Usage” data offloading strategy for deciding how many workloads need to be offloaded for each user;(2)a “Join the Best Queue” request routing strategy for balancing the system utility and queue backlogs for both mobile users and edge servers without prior knowledge of request arrivals;(3)a “Real-time Dynamic Adjustment” energy management strategy for controlling running states of edge servers to reduce power consumption of computing on edge services.Based on strategies above,DOMES realize the load balancing of two queue lengths and a [O(1/V);O(V)] tradeoff between the system’s utility and workloads with V.Simulations validate the effectiveness of DOMES in optimizing the system utility while balancing the double-queue workloads.