Research On Edge Computing Task Scheduling For Intelligent Buildings

Task scheduling is a hotspot in the field of edge computing.And intelligent building is an important application scenario.The essence of this problem is how to allocate lots of tasks to edge servers reasonably,that generated by building internal equipment.Thus,it is convenient to reduce the delay and task queuing caused by big data,as much as possible.The traditional cloud computing method has the disadvantages of high energy consumption,high delay,and poor security.It has become more and more difficult to adapt to the era of Internet of things.There is a growing trend to move data from the cloud to the edge side for nearby processing.In order to achieve efficient utilization of resources,it is important to make use of scheduling strategies and algorithms.The following research was to solve these problem.(1)The key technologies and user demand of intelligent building based on edge computing are researched,in this paper.Combining with the research status at home and abroad,the edge computing scheme of is determined.It is useful to overcome the disadvantages of high latency and high energy consumption in the traditional cloud computing mode,which deals with the mass tasks in intelligent buildings.(2)An edge computing task scheduling strategy for intelligent buildings is proposed.The tasks are split,then the compute-intensive tasks are offloaded to the cloud data center.In order to reduce the task queuing time,and improve the service response speed.(3)In the research of task scheduling,a scheduling scheme to minimize task completion time and energy consumption is proposed.The factors,such as communication transmission delay and task queuing time,were considered.Then,the genetic mutation mechanism of GA is introduced on the basis of SSA.SSA-GA was proposed.It not only makes up for the defect that SSA is easy to fall into local optimization,but also retains the advantages of few parameters and simple structure(4)A task scheduling scheme based on AHP model is proposed to provide users with better experience and satisfaction.On this basis,DAG task clustering is carried out on the tasks.The AHP model is improved,which greatly reduces the comparison times of the comparison matrix,further reduces the system delay and task processing time,and optimizes the scheduling scheme.