Design And Implementation Of Computation Offloading Modules For The IIoT

In recent years,Industrial Internet of Things(IIoT)has developed rapidly,and Internet of Things(IoT)devices have been widely used in the industry.In IIoT network architecture,edge computing and cloud computing are usually used to process the computing tasks generated by IIoT devices to solve the problem of insufficient computing power.However,edge computing and cloud computing also have limitations in data transmission,data storage,and resource allocation in the IIoT environment.Therefore,a reasonable computing offloading strategy is required to give full play to the advantages of edge computing and cloud computing.In response to the above problems,this paper proposes a computation offloading scheme for time and energy consumption in cloud-edge-terminal collaboration networks(COSTE).First of all,a cloud-side computing offloading model is established for the characteristics of multi-threaded processing and dynamic allocation of computing resources for servers.Furthermore,a fast unloading algorithm based on the restricted first adaptive algorithm and a precise unloading algorithm based on an improved genetic algorithm are proposed to unload computing tasks.Finally,an unloading algorithm for selected tasks based on a dynamic threshold-based unloading decision mechanism is proposed.In this paper,simulation experiments are carried out on IIoT devices and edge servers of different numbers and performances.Compared with particle swarm algorithm,COSTE has superior performance in reducing total consumption.Furthermore,in response to the needs of visual operation in IIoT,a computing offloading module for IIoT is designed and implemented based on the Django framework.First,it describes in detail the requirement analysis,structural framework and processing flow of the calculation unloading module.Then,the parameter configuration module,result display module,model and algorithm module and data storage module are designed and implemented.Finally,the module is tested in detail.The test results show that the designed computing offloading module can improve the convenience and efficiency of computing offloading,and meet the functional and performance requirements of industrial Internet of Things computing offloading.