Research On Offloading Strategy Of Edge Computing Task For Industrial Internet Of Things

With the continuous evolution of modern industry in the direction of intelligence,the number of industrial equipment has increased dramatically,and the demand for computing resources is increasing.However,due to the limitations of its own network architecture and related technologies,traditional industrial networks cannot meet the real-time requirements of computing resources and data in different industrial application scenarios.Edge computing flattens the originally concentrated resources and can provide localized computing resources with low delay and high reliability for industrial field equipment.As a new type of network architecture,Software Defined Network(SDN)can realize flexible scheduling of network data.Applied to the offloading of edge computing tasks in Industrial Internet of Things(IIoT),SDN technology can effectively ensure the real-time nature of the computing task data and meet different demands of different industrial applications for computing resources.In view of this,this thesis proposes an edge computing network architecture based on Software Defined Industrial Internet of Things(SDIIoT),and designs a method for offloading edge computing tasks.The main tasks are as follows:1.Through research and analysis of the key technologies of SDIIoT and edge computing,combining with the demand for real-time computing resources and data in industrial application scenarios,and referring to the industrial network architecture proposed at home and abroad,this thesis brings forward an edge computing network architecture based on SDIIoT.2.In order to reduce the transmission delay of computing task offloading,based on the network architecture proposed in this thesis,a task offloading transmission path algorithm based on link performance awareness is designed.The SDN controller perceives the network link performance parameters,analyzes the computing task request's demand for network resources,and reasonably plans the transmission path of the computing task offloading.3.By establishing the delay model for task offloading,using the maximum tolerance delay as the constraint condition,the primary selection rule is used to select the set of candidate computing nodes for the computing task offloading.Aiming at the decision problem of multi-parameter and multi-standard computing node selection,a method of offloading node selection for computing tasks based on fuzzy logic is proposed.4.The rationality of the transmission path algorithm of task offloading based on link performance perception and the selection method of the task offloading node based on fuzzy logic is verified in this thesis through the construction of an experimental simulation platform.The simulation results show that the task offloading transmission path algorithm based on link performance awareness proposed in this thesis can not only meet the requirements of different types of computing task flow for network performance parameters,but also select the transmission path with less delay and packet loss rate fluctuations.This thesis proposes a fuzzy logic-based method to select offloading nodes of computing tasks,which is superior to the compared algorithm in terms of the allocation of computing resources,the execution time of computing tasks and the number of unexecuted computing tasks.The research results of this thesis have an important reference value for edge computing task offloading under IIoT.