结合SDN控制器的TSN周期性带宽预留值计算方法

doi:10.11959/j.issn.1000-436x.2021204

摘要/Abstract

摘要：

近年来，随着 TSN 技术的发展，其低时延、低抖动、最大时延可预测的优势可以很好地应用于自动驾驶技术以及工业互联网。针对TSN协议中带宽预留技术预留至多75%（经验值）带宽的问题，同时结合SDN的数据面和控制面的分离技术，提出一种结合TSN-SDN控制器，在控制器层面规划整个拓扑的预留计划和路径的预留方案。该方案同时提供2种计算具体预留带宽值的方法，计算方法在每一跳按照时隙预留相应的带宽值。仿真环境实验结果表明，在达到时延要求的基础之上，很好地控制抖动为 0，同时还可以分时隙利用带宽，为多条流同时保障带宽利用率。

关键词: 时间敏感网络, 确定性网络, 带宽预留, 软件自定义控制器

Abstract:

In recent years, with the development of TSN, its low latency, low jitter and predictable maximum delay can be widely used in automatic driving technology and industrial Internet.To solve the 75% (empirical value) bandwidth problem of stream reservation technology in TSN protocols, and combine with the separation of data plane and control plane in SDN, a stream reservation solution was put forward, which was combined with TSN-SDN controller and scheduling the transmission path and reservation of the entire topology in the control plane.And two reservation computing methods was put forward, which would give out a bandwidth with time slot.The simulation experimental results show that it can control the jitter to 0 based on require latency, and also use the bandwidth based on time slot to guarantee the utilization of bandwidth.

Key words: time sensitive networking, deterministic network, stream reservation, SDN controller

中图分类号:

TP393

苏建忠, 张华宇, 朱海龙. 结合SDN控制器的TSN周期性带宽预留值计算方法[J]. 通信学报, 2021, 42(10): 23-31.

Jianzhong SU, Huayu ZHANG, Hailong ZHU. Computing method for periodic stream reservation in TSN combined with SDN controller[J]. Journal on Communications, 2021, 42(10): 23-31.

图/表 15

图1

表1

系统参数或公式"

参数或公式	含义
G(V,E)	网络拓扑中节点和链路集合
V	网络节点（终端、TSN talker、TSN listener、交换机等）
E	终端节点之间的连接链路（连接物理线路，如光纤、双绞线等）
$F^{[V_{a}, V_{b}]}, \forall f_{i}^{[V_{a}, V_{b}]} \in F^{[V_{a}, V_{b}]}$	节点a到b的帧流量
$α_{i}^{[V_{a}, V_{b}]} / B$	节点a到b的流量切分后对应第i帧大小
T/s	周期
β/s	流量的最大时延要求
$β^{[V_{a}, V_{b}]} / s$	节点a到b的整体时延
γ	数据流切分的帧数
δ/B	流量的最后一帧大小
$ε^{[V_{a}, V_{b}]} / (bit \cdot s^{- 1})$	节点a到b的链路总带宽值
$ω_{[b, e], i}^{[V_{a}, V_{b}]} / (bit \cdot s^{- 1})$	节点a到b的链路时隙预留带宽值，c和e
$[c, e] \in T, i \in F$	表示周期内起始时间，j 表示流量特征用于标识具体的一条流
$θ^{[V_{a}, V_{b}]} / s$	节点a到b信道时延
$ϑ_{V_{a}} / s$	节点a设备时延
$μ^{[V_{a}, V_{b}]} / s$	节点a到b流的链路传输时延
$ρ^{[V_{a}, V_{b}]} / s$	节点a到b流的最后一帧链路传输时延
$σ_{i}^{[V_{a}, V_{b}]} / s$	流的第i帧在节点a到b的发出时间
$τ^{[V_{a}, V_{b}]} / s$	节点a到b第一帧数据包到达时间
$φ^{[V_{a}, V_{b}]} / s$	节点a到b最后一帧数据包到达时间

表1

图2

表2

表3

表4

表5

表6

图3

图4

表7

表8

图5

图6

图7

参考文献 17

[1]	黄韬, 刘江, 汪硕 ,等. 未来网络技术与发展趋势综述[J]. 通信学报, 2021,42(1): 130-150.
	HUANG T , LIU J , WANG S ,et al. Survey of the future network technology and trend[J]. Journal on Communications, 2021,42(1): 130-150.
[2]	DUJOVNE D , WATTEYNE T , VILAJOSANA X ,et al. 6TiSCH:deterministic IP-enabled industrial Internet of things[J]. IEEE Communications Magazine, 2014,52(12): 36-41.
[3]	黄韬, 汪硕, 黄玉栋 ,等. 确定性网络研究综述[J]. 通信学报, 2019,40(6): 160-176.
	HUANG T , WANG S , HUANG Y D ,et al. Survey of the deterministic network[J]. Journal on Communications, 2019,40(6): 160-176.
[4]	JOHAS TEENER M D , GARNER G M . Overview and timing performance of IEEE 802.1AS[C]// Proceedings of 2008 IEEE International Symposium on Precision Clock Synchronization for Measurement,Control and Communication. Piscataway:IEEE Press, 2008: 49-53.
[5]	CRACIUNAS S S , OLIVER R S , CHMELíK M ,et al. Scheduling real-time communication in IEEE 802.1Qbv time sensitive networks[C]// Proceedings of Proceedings of the 24th International Conference on Real-Time Networks and Systems. New York:ACM Press, 2016: 1-8.
[6]	LAN/MAN Standards Committee of the IEEE Computer Society. IEEE Std 802.1QccTM-2018 Amendment 31:stream reservation protocol (SRP) enhancements and performance improvements[R]. 2018.
[7]	KUMAR R , HASAN M , PADHY S ,et al. End-to-end network delay guarantees for real-time systems using SDN[J]. arXiv Preprint,arXiv:1703.01641, 2017.
[8]	SCHRIEGEL S , KOBZAN T , JASPERNEITE J . Investigation on a distributed SDN control plane architecture for heterogeneous time sensitive networks[C]// Proceedings of 2018 14th IEEE International Workshop on Factory Communication Systems (WFCS). Piscataway:IEEE Press, 2018: 1-10.
[9]	吴文君, 姚海鹏, 黄韬 ,等. 未来网络与工业互联网发展综述[J]. 北京工业大学学报, 2017,43(2): 163-172.
	WU W J , YAO H P , HUANG T ,et al. Survey of development on future networks and industrial Internet[J]. Journal of Beijing University of Technology, 2017,43(2): 163-172.
[10]	王涛, 王永强, 王康 . 即时调度中周期调度最大化的带宽预留策略[J]. 软件, 2019,40(12): 118-123.
	WANG T , WANG Y Q , WANG K . Bandwidth reservation strategy for maximizing periodic scheduling in real-time scheduling[J]. Computer Engineering ＆ Software, 2019,40(12): 118-123.
[11]	陈赓, 夏玮玮, 沈连丰 . 基于多门限预留机制的自适应带宽分配算法[J]. 通信学报, 2014,35(12): 78-88.
	CHEN G , XIA W W , SHEN L F . Adaptive bandwidth allocation algorithm based on the multi-threshold reservation mechanism[J]. Journal on Communications, 2014,35(12): 78-88.
[12]	LIN Y Y , WU Q S . Complexity analysis and algorithm design for advance bandwidth scheduling in dedicated networks[J]. IEEE/ACM Transactions on Networking, 2013,21(1): 14-27.
[13]	WANG T , WU C Q , WANG Y Q ,et al. Multi-path routing for maximum bandwidth with K edge-disjoint paths[C]// Proceedings of 2018 14th International Wireless Communications ＆ Mobile Computing Conference (IWCMC). Piscataway:IEEE Press, 2018: 1178-1183.
[14]	NAYAK N G , DüRR F , ROTHERMEL K . Time-sensitive software-defined network (TSSDN) for real-time applications[C]// Proceedings of Proceedings of the 24th International Conference on Real-Time Networks and Systems. New York:ACM Press, 2016: 193-202.
[15]	BHM M , OHMS J , KUMAR M ,et al. Dynamic real-time stream reservation for IEEE 802.1 time-sensitive networks with OpenFlow[C]// 8th International Conference on Applied Innovation in IT.[S.n.:s.l], 2020: 1-8.
[16]	POZO F , STEINER W , RODRIGUEZ-NAVAS G ,et al. A decomposition approach for SMT-based schedule synthesis for time-triggered networks[C]// Proceedings of 2015 IEEE 20th Conference on Emerging Technologies ＆ Factory Automation (ETFA). Piscataway:IEEE Press, 2015: 1-8.
[17]	BJ?RNER N , PHAN A D , FLECKENSTEIN L . νZ - an optimizing SMT solver[C]// Tools and Algorithms for the Construction and Analysis of Systems. Berlin:Springer, 2015: 194-199.

名称	设备用例	周期	数据大小
LiDAR₁～LiDAR₅	激光雷达	300 μs	1 200 KB
MMWR₁～MMWR₂	毫米波雷达	50 ms	100 KB
ULR₁～ULR₂	超声波雷达	1 ms	500 B
HDcamera₁～HDcamera₄	高清摄像头	33 ms	26.5 MB
SVcamera₁～SVcamera₄	环视摄像头	40 ms	5 MB

设备名称	平均时延/s	时延标准偏差/s
LiDAR₁	9.989 96×10^-5	3.255×10^-6
LiDAR₂	0.000 12	0
LiDAR₃	0.000 100 608	3.3395 3×10^-6
SVcamera₂	0.004 124 02	1.229 71×10^-5
SVcamera₃	0.004 123 94	1.229 74×10^-5
HDcamera₂	0.021 785 3	4.659 28×10^-6
HDcamera₃	0.021 785 1	4.666 54×10^-6

设备名称	平均时延/s	时延标准偏差/s
LiDAR₁	9.991 39×10^-5	3.301 65×10^-6
LiDAR₂	0.000 120 002	0
LiDAR₃	0.000 100 623	3.382 04×10^-6
SVcamera₂	0.004 123 87	1.234 82×10^-5
SVcamera₃	0.004 123 72	1.235 6×10^-5
HDcamera₂	0.021 785 5	4.696 8×10^-6
HDcamera₃	0.021 785 3	4.640 21×10^-6

设备名称	平均时延/s	时延标准偏差/s
LiDAR₁	9.949 87×10^-5	2.055 74×10^-7
LiDAR₂	0.000 100 445	3.402 17×10^-7
LiDAR₃	0.000 100 05	4.684 53×10^-7
SVcamera₂	0.004 970 96	0
SVcamera₃	0.004 111 48	4.556 79×10^-7
HDcamera₂	0.021 784 2	8.499 88×10^-8
HDcamera₃	0.021 784 1	2.386 02×10^-7

设备名称	平均时延	时延标准偏差
LiDAR₁	0.000 105 487	2.444 96×10^-5
LiDAR₂	0.000 106 44	2.457 46×10^-5
LiDAR₃	0.000 106 036	2.459 97×10^-5
SVcamera₂	0.004 980 336	0
SVcamera₃	0.004 210 84	7.756 47×10^-5
HDcamera₂	0.021 793 2	3.192 85×10^-5
HDcamera₃	0.021 793 1	3.206 59×10^-5