Research On Service Function Chain Orchestration And Migration For Industrial Internet Of Things

The Industrial Internet of Things(IIoT)integrates the Internet of Things with industry.It facilitates the interconnection of sensors,production lines,and various smart devices in modern industrial production and automation environments.In the IIoT environment,various devices will generate a large number of data streams.How to process the forwarding of these data streams and reduce network latency has become a major challenge for the IIoT.At the same time,devices in the IIoT typically move frequently,such as mobile robots,intelligent robotic arms,driverless trucks,and mobile communication devices carried by workers,which also places new demands on the flexibility offered by IIoT services.In order to solve the real-time,resource diversity,security and other requirements in the Industrial Internet of Things,many researches have introduced Network Function Virtualization(NFV)and Mobile Edge Computing(MEC),and the service function chain(The orderly arrangement of Service Function Chain(SFC)in the edge network can make the data forwarding in the Industrial Internet of Things more flexible,meet the QoS requirements of users,and greatly improve the manageability and flexibility of the Industrial Internet of Things.However,most of the current service function chain orchestration research does not consider the situation that the orchestration results may be untrustworthy caused by the sharing of network resources by different vendors in the edge-center environment,and the efficiency of service orchestration needs to be further improved.In the scenario of frequent device movement in the IIoT,the deployment of fixed service function chains will bring long service delays.However,most of the current research does not consider the impact of user mobility,but only considers passive migration scenarios due to node overload.In view of the shortcomings of the current service function chain orchestration and migration algorithm,this paper conducts researches in the following three aspects:(1)In order to further improve the efficiency of orchestration and improve the reliability of service orchestration,this paper proposes a federated service function chain orchestration algorithm based on joint optimization of energy consumption and delay.The algorithm in this paper firstly combines federated learning with traditional reinforcement learning,which accelerates the convergence of the model and enhances the robustness of the model.At the same time,this paper introduces reputation theory to evaluate the reliability of different orchestration models,improve the anti-interference ability of federated learning,and improve the accuracy of orchestration results.The simulation results in this paper show that the proposed algorithm can meet the effects of accelerating convergence and optimizing energy consumption and delay.(2)In order to better adapt to the dynamic movement of devices in the Industrial Internet of Things,this paper proposes a service function chain migration strategy based on device mobility and network status.In the migration algorithm proposed in this paper,the position data collected by GPS is first calibrated,the LTSM model is used to model the position of the device at the next moment,and the prediction result combined with the Lyapunov optimization algorithm is used to make a decision on the migration of SFC.The simulation shows that the algorithm has achieved certain effects in reducing the probability of service interruption,reducing the delay and controlling the migration cost.(3)Based on the algorithm mentioned above,this paper designs and implements a service function orchestration and migration simulation platform.This paper firstly designs and sorts out the functions of the platform,and builds the basic framework and underlying database.After that,this paper further designs the service chain orchestration module,the service function chain migration module and the topology display module.Technically,the MVC architecture is adopted as a whole,and various Spring frameworks are used for integration.The front-end interface is implemented with the LayUI framework,and Vis.js is used for topology drawing.The back-end development adopts the Spring Boot framework,which uses its inversion control and dependency injection,etc.Features simplify the development process.The database uses MySql.Finally,the system is tested,and the test results show that this paper can complete the expected functions and help the operation and maintenance personnel to arrange and simulate the service function chain.