Research On Traffic Scheduling Mechanisms Of Time-Sensitive Networking

With the rapid development of Internet technology and 5G technology,more and more emerging industry applications require bounded low-latency,low-jitter,and high-reliability network transmissions to support application operations,such as industrial control,5G mobile fronthaul,and telemedicine Etc.It is difficult for traditional networks that provide best-effort services to meet application requirements.Although industrial Ethernet can provide real-time communication services,there are many standards and it is difficult to be compatible and interconnected.In order to solve the real-time synchronous transmission of audio and video streams,the IEEE 802 Ethernet Audio / Video bridging standard working group uses precision clock synchronization based on traditional Ethernet to reduce transmission delays by ensuring bandwidth.Audio and video applications provide high-quality transmission services.In order to extend Ethernet technology to industrial,automotive and other fields based on the Ethernet AVB technology standard,IEEE 802.1 established a Time-Sensitive Networking(TSN)working group to enhance Ethernet-related functions.The current TSN standardization areas include flow classification,flow shaping,flow scheduling,and frame preemption.In time-sensitive networks,traffic scheduling is one of the key technologies for TSN to achieve low latency and low jitter performance.Therefore,the IEEE 802.1 working group provides related traffic scheduling algorithms and standards to achieve related functions,such as 802.1Qav based credit value shaping algorithm,802.1Qbv gated scheduling algorithm,802.3br and 802.1Qbu frame preemption algorithms,802.1Qch Circular queue forwarding algorithm,802.1Qcr asynchronous shaping algorithm.However,the current TSN traffic scheduling mechanism cannot meet the traffic performance requirements in specific scenarios,and researchers need to design more efficient traffic scheduling mechanisms based on application scenarios.In order to solve the delay optimization problem of path selection of industrial time triggered flow,this paper proposes a delay-aware differential multipath forwarding mechanism for industrial time triggered flow.The differential multipath forwarding mechanism proposed in this paper first analyzes the end-to-end delay,designs a multipath forwarding index based on the transmission delay and processing delay,and then converts the multipath forwarding problem into a 0-1 integer linear programming problem.Finally,through the differential multipath forwarding algorithm,the flow forwarding path set is solved,and the low-latency parallel transmission of the flow is realized.Simulation results show that,compared with load-balanced multipath forwarding and Hop-Load forwarding mechanisms,the scheme reduces the end-to-end average time delay of time-triggered flows by 22.55% and 17.32%,respectively,and satisfies the deterministic flow of low-latency transmission demand.In order to solve the problems of background flow in mobile fronthaul networks that can affect in-phase and positive communication,resulting in low jitter and low link throughput,this paper proposes a HC-TAS(Hierarchical Cross-TAS)-based traffic Scheduling mechanism.This mechanism uses a crossed two-level queue scheduling mechanism to maximize the use of time slots without introducing new jitter due to background flow.The HC-TAS mechanism proposed in this paper is based on the traditional TAS mechanism.On this model,an input-queue scheduling algorithm based on frame length matching and an out queue scheduling algorithm based on queue status are proposed,so that smaller data frames are not blocked by large data frames.To improve the utilization of time slots in the scheduling process.The simulation results show that: under the condition that the jitter is guaranteed to be 0,compared with the typical TAS scheduling scheme,this scheme improves the throughput of the link by18.7%,and meets the requirements of low jitter and high bandwidth utilization of 5G mobile pre-streaming..This dissertation studies the optimization of delay in time-triggered flow scheduling in industrial networks and background flow scheduling in mobile fronthaul networks.A delay-aware differential multipath forwarding mechanism for industrial time-triggered flow and a hierarchical cross-flow scheduling mechanism for time-aware shaper for mobile forward-flow are proposed.Improved network latency,jitter,and throughput performance.The research results will provide a technical basis for the application and development of TSN.