Research On Task Scheduling Strategy In 5G Mobile Edge Computing

With the widespread popularization of mobile devices such as smart phones and the rapid development of Internet of Things(Io T),various kinds of new applications for mobile devices and Io T are emerging,which results in an explosive growth of data traffic in mobile networks.To satisfy the performance requirements of these applications,the 5th generation mobile communication technology(5G)is proposed,which can bring the higher data traffic and better experience of service.As a key technology in 5G networks,mobile edge computing(MEC)deploys the computing and storage resources at the network edge,which can provide users with the low-latency and high-reliability services.In the field of 5G mobile edge computing,the problems of task offloading for mobile devices,task scheduling in edge clouds and collaborative task scheduling between mobile devices and the edge cloud are the hot topics,which play the key roles in improving users' experience of service and enhancing system performance.Firstly,we study the problem of task offloading for mobile devices.By considering both the burst in task arrival and randomness in wireless channel condition,we construct a dynamic task offloading model,and formulate a stochastic optimization problem with the goal of minimizing the transmission energy consumption while guaranteeing the queueing delay of tasks.However,it is extremely difficult to directly solve this stochastic optimization problem.To address this problem,by applying Lyapunov optimization techniques,we transform the original problem into an one-dimensional linear knapsack problem.Then,we propose an energy efficient dynamic offloading algorithm(EEDOA)to solve this transformed problem,and also analyze the performance of EEDOA mathematically.Simulation results show that EEDOA can reduce the transmission energy consumption by about30%.Secondly,we investigate the problem of task scheduling in the edge cloud.By analyzing the limitation and heterogeneity for the resources in MEC servers,we develop a task scheduling model for delay-sensitive tasks.An optimization problem is formulated,which minimizes the running cost of all MEC servers while satisfying the delay requirement of each task.Then,we prove that the proposed optimizationproblem is NP-hard.To solve this NP-hard problem effectively,we propose a heuristic task scheduling algorithm based on best-fit greedy strategy.Simulation results show that the approximate ratio between the solution of our proposed algorithm and the optimal solution is less than 1.2 for 95% of the data sets we use.Finally,we study the problem of collaborative task scheduling in MEC.To provide mobile devices with the sustainable energy supply,we incorporate wireless power transfer(WPT)technologies into MEC.By considering the coupling between WPT and task scheduling,and randomness in system environments,we establish a WPT-based collaborative task scheduling model.With the constraint of queue stability,we formulate an optimization problem to minimize the energy consumption of MEC system.To make the formulated optimization problem more tractable,we transform it into a convex optimization problem by using Lyapunov optimization methods.Then,with the help of Lagrange dual method,we design an iterative collaborative task scheduling algorithm.Simulation results validate the effectiveness of the proposed algorithm in reducing energy consumption.