Research On Optimal Scheduling Of End-Edge-Cloud Collaborative Resources Based On Task Characteristics

Cloud computing centers have large-scale resources required to perform tasks.However,with the rise of Internet of Things technology,the popularization of wireless networks,the ever-increasing volume of intelligent terminals and sensor data,the real-time and diversification of user service requirements,the demand f or data communication and computing has surged,the traditional computing mode that uploads all tasks to the cloud is no longer suitable for the real-time processing requirements of tasks,and the utilization rate of platform resources decreases,which makes cloud computing face severe challenges.The emerging architecture of edge computing can provide certain computing,storage,network and communication capabilities,and has the characteristics of distributed,low latency,and high service quality,effectively making up for the delay of cloud computing processing tasks.The capacity is much lower than that of cloud nodes.The selection of an efficient resource scheduling optimization framework and scheduling algorithm is directly related to the computing service quality of the end-to-end cloud system,which can effectively overcome resource scheduling problems caused by factors such as resource dispersion and demand diversification,achieve the goal of resource scheduling optimization,and provide safe,reliable and efficient computing services.In view of the complementary characteristics of cloud-edge computing,this paper proposes a consistent optimization framework for device-edge-cloud collaborative resource scheduling based on task characteristics.Ac cording to task characteristics and service requirements,tasks are divided into two types:computing-intensive and delay-sensitive.Cooperative mode and scheduling strategy processing.Aiming at the characteristics of high computing demand for computationally intensive tasks,the edge-cloud collaborative scheduling is selected,and the optimization goal is determined to reduce system load and energy consumption,and a resource scheduling strategy based on optimized particle swarm optimization is proposed.The swarm algorithm schedules sub-tasks to form cloud decisions,which are offloaded to edge nodes for execution through computing offloading technology,and the edge layer feeds back resource usage to the cloud in real time to achieve cloud-edge closed-loop collaborative control.Through MATLAB simulation experiments,the optimization strategy and traditional particle swarm algorithm Compared with the MGA-PSO algorithm,the effect is obvious in reducing the energy consumption of the system,and with the increase of the number of subtasks,the effect is more significant.Aiming at the characteristics of delay-sensitive tasks that have high requirements for response delay,end-edge collaborative scheduling is selected,and the optimization goal is to reduce t he task response delay while maintaining the system load as much as possible.A resource scheduling strategy based on improved ant colony algorithm is proposed.The MATLAB simulation experiment compares the improved ant colony algorithm,the traditional ant colony algorithm the first-come-first-served algorithm and AC-FSL algorithm.The results show the feasibility of the scheduling strategy.The resource scheduling strategy can effectively shorten the parallel processing time of tasks in the edge computing environment and reduce the Edge computing environment load to maintain system load balance.