The Design And Implementation Of TSN Traffic Scheduling Method Based On Flexible Time Window

To meet the requirements of industrial Internet for real-time performance and reliability,Ethernet introduces time-sensitive network(TSN)technology to meet the requirements of industrial Internet for efficient communication.Traffic scheduling is one of the important mechanisms of time-sensitive networks.Scheduling algorithms are used to schedule and control different traffic in terms of priority,bandwidth and delay,so as to ensure the real-time,reliability and stability of networks.Most of the existing traffic scheduling works are based on fixed routing schemes,which are difficult to respond to the burst slot contention,and do not consider the influence of non-periodic traffic on scheduling.In addition,due to the differences in device iteration in the actual network,clock synchronization in the network is difficult to achieve,and most current traffic scheduling algorithms are based on the premise of clock synchronization,the calculation results can not meet the needs of the actual industrial Internet.Aiming at the above two problems,this paper studies the design and implementation of TSN traffic scheduling method based on flexible time window,proposes a comprehensive routing and GCL scheduling problem model and a scheduling strategy oriented to differentiated traffic,and designs a traffic scheduling algorithm based on tabu search.Implement the mixed scheduling of periodic and aperiodic traffic in Ethernet.In order to solve the problem of clock unsynchronization between devices in TSN network,this paper implements a flexible windowing based GCL scheduling,which realizes the traffic scheduling in non-synchronous TSN network at the cost of a small maximum end-to-end delay and bandwidth utilization.Finally,the TSN traffic scheduling simulator is designed and implemented,and the scheduling scheme is calculated by the traffic scheduling algorithm.The NS3 module is used to simulate the TSN traffic scheduling based on the input network topology.