SDN-based Industrial Internet Traffic Scheduling On Fog Computing Platform And Algorithm Implementation

As the Industrial Internet of Things continues to grow in popularity,a large number of machines,goods,control systems and information systems will be interconnected to the Industrial Internet.As a result,when an emergency occurs,traffic will increase significantly and the network will become congested.In the worst case scenario,machines may not work properly.Today an increasing number of smart end devices and infrastructures have formed intelligent systems.And these systems generate time-sensitive data(e.g.,control data,fault monitoring data,etc.)that is transmitted within strict time and reliability rules.Primitive standard Ethernet meets bandwidth requirements while maintaining scalability and cost effectiveness,but is not suitable for applications where real-time performance is critical.In the Industrial Internet,physical processes are controlled through a distributed set of sensors,actuators and controllers.Cyber-Physical Systems(CPS)rely on computer networks to transmit sensor data and actuator commands to the controller,respectively.Typically,such CPSs are time-sensitive systems in which network delays(including those caused by packet loss)and jitter can affect the quality of control of the CPS.When two consecutive packets are lost,machines on the production floor may malfunction.Due to the complexity of industrial communication and automation,studying and improving the latency performance of networks is important for the future development of industrial communication and automation control.Load balancing of network links and network scheduling algorithms are very important parts of the network.In this thesis,based on the problem that traditional distributed network links have low bandwidth utilization and are not easily scheduled,the software-defined network(SDN)network architecture is utilized to obtain global network information,i.e.,the network state information of each SDN switch is extracted by the SDN controller through centralized control.This information is allocated by the fog computing server through network utility maximization using a weighted maximum-minimum fair algorithm for bandwidth allocation.A novel bandwidth allocation algorithm is optimized in terms of reducing the bandwidth allocation time,which is then used to reduce the network latency by embedding the bandwidth allocation into the traffic scheduling algorithm.The experimental results show that the bandwidth allocation algorithm proposed in this paper can effectively reduce the bandwidth allocation convergence time.And the scheduling algorithm proposed in this thesis using the results of bandwidth allocation can effectively reduce the network delay,and its results are better than the classical scheduling algorithm.Therefore,it can be used to solve the scheduling problem of industrial Internet.