Research On Task Scheduling In Industrial Intelligence Edge Computing

In recent years,with the continuous development of information and communication technologies,the number of mobile devices and data traffic have increased significantly.Because the traditional cloud computing framework is deployed in the cloud away from the user,there are problems such as high latency,low bandwidth,and privacy security threats,which cannot meet the requirements of low latency,high bandwidth,and privacy security of the new technology.As an emerging architecture,edge computing deploys some cloud computing services to the edge of the network close to users through edge servers,and solves the problems existing in the current cloud computing framework,and has become one of the current research hotspots.In the traditional edge computing framework,the task scheduling problem has been widely concerned,and the related theories,techniques and algorithms are also relatively mature.However,the research on task scheduling in the industrial intelligence edge computing framework is not complete enough.Consider the limited computing resources and storage resources of the edge server,the high time sensitivity of computing and offloading tasks,and the high security requirements for data privacy.Therefore,it is a significant work to study the industrial intelligence edge computing task scheduling problem.Based on the analysis of the research status of scheduling algorithms and incentive mechanisms in edge computing at domestic and foreign,this paper firstly proposes an industrial intelligence edge computing task scheduling problem model,and proposes non-cooperative non-lag task scheduling and cooperative lag task scheduling based on the model,and then designs the corresponding scheduling algorithm and incentive mechanism for the two application scenarios.Finally,the feasibility of the designed scheduling algorithm and incentive mechanism is proved by theoretical proof and simulation experiments.The main research contents and innovative work of this paper include:(1)A model of industrial intelligence edge computing task scheduling problem is proposed.Firstly,the in-depth analysis of industrial intelligence edge computing task scheduling problem is carried out.According to the characteristics of the problem,two application scenarios of non-cooperative non-lag task scheduling and cooperative lag task scheduling are proposed.Secondly,the design of industrial intelligence edge computing task scheduling problem is described.The necessity of incentive mechanism finally realized the model of industrial intelligence edge computing task scheduling problem.(2)Scheduling algorithm and incentive mechanism are designed for non-cooperative non-lag task scheduling scenarios.Firstly,the Edge Server Selection(ESS)algorithm is designed to meet the bandwidth and privacy security requirements.Secondly,the task scheduling problem in this scenario is an NP-hard problem,and an adaptive non-lag algorithm based on greedy algorithm is designed.The Adactive Non-Lag Scheduling(ANS)algorithm is used to achieve the goal of efficient use of edge servers.Third,an incentive mechanism satisfying personal rationality and quotation authenticity is designed.Finally,the proposed scheduling algorithm is proved by theoretical proof and simulation experiments.(3)A scheduling algorithm and incentive mechanism are designed for the cooperative lag task scheduling scenario.Firstly,a Minimize Total Lag Time(MTLT)algorithm based on dynamic programming is designed,and the optimality of the algorithm is proved.Secondly,a reliable solution is proposed for the performance problem of the MTLT algorithm under large tasks.Based on the algorithm,an adaptive Lag Scheduling(ALS)algorithm is proposed.Third,an incentive mechanism for satisfying personal rationality and quotation authenticity is designed again for ALS algorithm.Finally,the proposed scheduling algorithm is proved by theoretical proof and simulation experiments.