Research On Task Offloading And Resource Allocation Algorithm Based On Mobile Edge Computing

With the popularity of smart mobile devices and the advent of th e Internet of Things(IoT),the amount of data of user requests is growing rapidly.The centralized processing mode represented by cloud computing can no longer meet the high real-time requirements of tasks.Mobile edge computing(MEC)emerged as a computing model that provides services near the data generation source.However,the computing capability and resources of the MEC server are insufficient to process the long-cycle intensive task data,so MEC and cloud computing complement and optimize each other to form the cloud-edge collaborative computing model.Whether in the MEC environment or cloud-edge collaborative computing environment,how to efficiently offload tasks to suitable computing nodes for processing and better allocate various resources is an important research direction in this field.In this paper,we research the problem of task offloading and resource allocation in the MEC environment and cloud-edge collaborative computing environment,respectively,and proposed an effective solution based on the deep reinforcement learning method.The purpose of the research is to reduce task response time and energy consumption.The specific research work is summarized as follows:(1)An adaptive algorithm based on deep Q-learning is proposed to solve the problem of task offloading and resource allocation in the MEC environment.In the environment,there are multiple user equipment and multiple MEC server distribution points.The user equipment has mobility and simulates task generation in the form of a Poisson distribution.The proposed algorithm has self-learning ability,and it can determine whether the task needs to be offloaded and assign a suitable computing node for the task,continuously learns during the algorithm training process to improve decision accuracy.Compared with other comparison algorithms,the proposed algorithm has the best performance in reducing the average task response time and total system energy consumption,improving the system utility,which can meet the profits of users and service providers.(2)A task offloading and resource allocation algorithm under resource and reliability constraints is proposed for the cloud-edge collaborative computing environment.Compared with the MEC model,the cloud-edge collaborative computing model adds a cloud computing layer.Therefore,the task data generated by user devices can be calculated in three types of computing methods: local device,offload to the MEC server or cloud server.In addition to the constraints of computing capabilities and resources in mobile edge computing,the reliability of computing nodes is also considered in the cloud-edge collaborative computing environment.An online learning algorithm based on deep reinforcement learning is proposed for the objective optimization problem.Through the evaluation and verification of simulation experiments,the proposed algorithm can effectively improve the quality of experience of users and reduce system energy consumption.