Research On Routing And Scheduling Algorithms For The Simultaneous Transmission Of Diverse Data Streaming Services In The Industrial Internet

With the development of industrial Internet technologies such as big data and edge computing,information technology(IT)and operation technology(OT)has been accelerated,and the demand for real-time and deterministic industrial field equipment has increased.OLE for Process Control Unified Architecture(OPC UA)is a new standard of the OPC Foundation at the application layer.It provides a service-oriented architecture(SOA)for industrial equipment that conforms to the OPC UA standard by defining abstract services.Data Access(DA),History Data Access(HDA)and Alarm and Event(A?E)provide interconnection,interoperability and semantic interoperability for data transmission of industrial field devices.As a standard of the data link layer,Time Sensitive Network(TSN)provides a set of IEEE802.1 protocol suites to provide clock synchronisation,resource reservation and deterministic scheduling for data transmission.Reliable,low-latency data communication.However,there is still a need for network connectivity,data connectivity,and business connectivity in the current production process of intelligent manufacturing on industrial sites.Researching the integration of OPC UA and TSN can effectively solve the problem of time insensitivity and the inability to transmit OPC UA data.Support highreliability synchronous real-time data transmission and other issues.The OPC UA Over TSN research mainly focuses on data mapping,routing,and scheduling.Data mapping generally adopts a five-layer mapping method: application layer-transport layer-network layer-data link layer-physical layer.Although it has a relatively convenient mapping interface,the unpacking and encapsulation process in the mapping process will generate a large delay,and its reliability cannot be guaranteed for the transmission of a type of synchronous real-time data.In addition,the swarm intelligence optimisation algorithm is often used in the process of routing and scheduling,and the method of determining the scheduling table of TSN switches by manual parameter adjustment and optimisation of fitness function solution still does not have relatively perfect reliability.The optimisation and adjustment of its parameters do not meet a specific requirement: network-like expansion and other issues.Based on the above two types of problems,this paper proposes a multi-service flow hybrid mapping method for data mapping.By directly mapping synchronous realtime data from the application layer to the data link layer,and using Pub/Sub communication mode,while retaining asynchronous The five-layer mapping method of real-time and non-real-time data uses TCP and UDP communication methods respectively,thereby effectively ensuring the transmission reliability of synchronous real-time data streams.In addition,for the process of routing and scheduling,an RSTSN algorithm based on data preprocessing is proposed.Among them,the data flow preprocessing adopts the K-means clustering algorithm to classify the hybrid data streams,and the network topology generated by the BA model is used as the input of the RSTSN algorithm.The RSTSN algorithm adopts the Z3 solver based on Satisfiability Modulo Theories(SMT),and by adding the constraints of data stream transmission,including data stream constraints,port constraints and delay constraints,etc.,to solve the Gate Control List(GCL)that satisfies routing and scheduling)to facilitate the design of Time Aware sharper(TAS)that meets scheduling needs.This paper compares and evaluates the RSTSN algorithm,the ant colony and the improved ant colony algorithm from the perspective of the network topology scale and the number of data streams.The relationship between the scheduling performance and the network topology scale and the number of data streams is comprehensively analyzed,as well as the comparison of the running time of the algorithm itself.The experimental results show that the RSTSN algorithm proposed in this paper has better average transmission delay and jitter,and effectively improves the scheduling success rate.Compared with the improved ant colony algorithm,the RSTSN algorithm can significantly reduce the scheduling calculation time in the case of large-scale network topology.