Optimization And Analysis Of Scheduling Mechanism In Time-Sensitive Networking

Time-Sensitive Networking(TSN)provides real-time system with a universal solution,which guarantee end-to-end transmission with bounded low latency for time-sensitive time-triggered flows based on the existing mechanisms.At present,the IEEE 802.1 working group has defined a series of standards for it,and various mechanisms have gradually matured.However,there are still some problems to be solved and room for optimization of scheduling shaping mechanismwhich is core of TSN in the actual applications.There is also time-critical event-triggered flows in some systems such as advanced driver assistance systems,aircraft,and industrial automation.The currently widely used TSN based on Timeaware Shaper(TAS)mechanism only considers periodic time-triggered flows and cannot be effectively compatible with unpredictable eventtriggered flows,and asynchronous scheduling cannot be applied due to the heavy load of the mixture of time-triggered and event-triggered flows.In addition,the rapidly developing cyclic queue forwarding(CQF)retains the forwarding controllability of the gate scheduling in TAS,and at the same time reduces the complexity of the gate control lists.However,the length of the queues directly affects the network performance as a key parameter.The length of the output port cache queues directly affects the performance of network scheduling and is constrained by hardware resources in implementation.This paper proposes a shaping and scheduling method that is capable of accommodating emergency event-triggered flows to address the limitations of TAS mechanisms in real scenarios.The method extends the TAS mechanism in the data forwarding plane and designs a scheduling scheme.It achieves the scheduling of burst event flows by periodically reserving transmission resources for event-triggered flows and improving resource utilization through dynamic priority and resource reuse strategies.Experimental results indicate that this method can simultaneously meet the scheduling requirements of both sudden event-triggered flows and timetriggered flows,effectively reduce the transmission delay of eventtriggered flows,and improve the utilization of network bandwidth.To address the issues of performance analysis and optimization as well as the selection of key parameters in the practical application of CQF mechanisms,this paper proposes a network calculus-based performance analysis method for CQF.This method abstracts network nodes and flows and constructs queue service and data flow models for the CQF mechanism,enabling the analysis of network system service performance and determination of the best queue parameters to optimize performance.Simulation experiments demonstrate that this method can obtain reasonable and feasible queue length values with a small amount of computing resources,providing a reliable performance analysis tool for scheduling transmission based on CQF and improving the efficiency of scheduling computation.