Online Scheduling Algorithm Research Of Real-time Streams In Time-sensitive Networking

Time-Sensitive Networking(TSN)is an industrial network technology that is currently being actively promoted by the industry.It is one of the most promising technologies for ensuring real-time and deterministic transmission of information in industrial production.In the industrial Internet applications,such as automatic drive,mobile robot navigation,equipment remote operation and maintenance,and predictive maintenance require network devices to interact with a dynamic and uncertain production environment in real-time manner.And industrial network stream scheduling needs the ability to adapt to the dynamic environment.Traditional TSN technology uses static scheduling strategies to realize real-time scheduling and transmission of network streams.Once the production environment changes,the original static scheduling strategies may completely fail and require complete reconstructions.How to design a safe and efficient real-time stream dynamic scheduling strategy is an open problem that needs to be solved urgently in the field of TSN.Therefore,this article will focus on online scheduling algorithm for two types of real-time streams in TSN.On the one hand,for the Time-Triggered(TT)real-time stream,a hard realtime dynamic scheduling framework with schedulability guarantee is proposed.On the other hand,for the Event-Triggered(ET)real-time stream,a soft real-time heuristic scheduling algorithm and max-sum algorithm are proposed to guarantee low latency.The specific work is as follows:(1)Aiming at the problem of real-time transmission of periodic TT streams in TSN,a realtime scheduling algorithm framework based on Virtual Earliest Deadline First(VEDF)is proposed.In the framework,each TT stream corresponds to an active interval on each switch through which it flows.Only when the TT stream is in the corresponding active interval,it participates in the priority competition according to the VEDF algorithm;otherwise,it is scheduled according to the lowest priority.In this article,an iterative algorithm for rationally configuring the active interval of TT streams is given.The sufficient condition for the proposed algorithm to successfully schedule TT streams is proved theoretically,and the schedulability of TT stream is guaranteed.Experiments show that the VEDF algorithm has excellent performance in the aspects of successful scheduling ratio and runtime,and the runtime is significantly reduced while maintaining an acceptable ratio of successful scheduling.(2)Aiming at the problem of real-time transmission of non-periodic ET streams in TSN,heuristic scheduling algorithm and max-sum algorithm are proposed.The problem is attributed to the problem of Job-Shop scheduling with soft real-time constraints.This article proposes a heuristic algorithm to set the scheduling priority of ET streams,not only taking into account the impact of the priority of the ET stream on its own delay,but also considering the impact of the ET stream priority on the delay of other ET streams.The max-sum algorithm takes into account competition on different switches,and uses the start time of the transmission operations as variable nodes and the constraint functions between the transmission operations as function nodes.The variable nodes and the function nodes pass messages iteratively to calculate the start time of the transmission operations.Experiments show that the heuristic algorithm has better performance in reducing end-to-end delay than most priority rule methods.Compared with the genetic algorithm,the max-sum algorithm has little difference in transmission completion time when the number of streams is small.(3)Aiming at the application problems of the above three algorithms in TSN,the possibility of applying them to TSN gate control list configuration protocol is discussed,which provides a new method for the dynamic scheduling and transmission of TSN real-time network streams.