BBR拥塞控制算法延迟及带宽探测优化

计算机与现代化 ›› 2022, Vol. 0 ›› Issue (10): 113-120.

BBR拥塞控制算法延迟及带宽探测优化

(1.中国科学院声学研究所国家网络新媒体工程技术研究中心，北京100190;2.中国科学院大学，北京100049)

出版日期:2022-10-20 发布日期:2022-10-24
作者简介:黄宏平(1997—)，男，江西上饶人，硕士研究生，研究方向：信息中心网络，E-mail: huanghp@dsp.ac.cn; 通信作者：朱小勇(1982—)，男，安徽怀宁人，研究员，博士，研究方向：信息中心网络，E-mail: zhuxy@dsp.ac.cn; 王志远(1995—)，男，河北衡水人，博士研究生，研究方向：信息中心网络，传输协议，E-mail: wangzy@dsp.ac.cn。
基金资助:
中国科学院战略性先导科技专项课题(XDC02070100)

Improving Latency and Bandwidth Probe of BBR Congestion Control Algorithm

(1. National Network New Media Engineering Research Center, Institute of Acoustics, Chinese Academy of Sciences,
Beijing 100190, China; 2. University of Chinese Academy of Sciences, Beijing 100049, China)

Online:2022-10-20 Published:2022-10-24

摘要/Abstract

摘要： 传统基于丢包的拥塞控制算法因为其高丢包率和引发缓冲区膨胀问题已经不能满足许多应用对网络性能的要求。谷歌提出的BBR（Bottleneck Bandwidth and Round Trip）算法以其抗丢包、高带宽利用率和低延迟等特性受到广泛关注与研究。但是BBR还存在排队延迟仍然较高、在RTT（Round Trip Time）较小环境下表现不佳、带宽探测不及时等问题。本文对BBR排队延迟和收敛性进行分析，进而提出改进方法：限制在外数据包数，并根据网络反馈适时减少拥塞窗口大小来降低延迟；在RTT较小环境下，将探测RTT阶段之前的带宽估计延续到探测RTT阶段之后；设置平稳状态最长保持时间及时退出平稳周期并进入探测周期。在NS3中的仿真实验结果表明，改进BBR降低了RTT及其抖动，提高了算法的收敛速度；能够在RTT较小环境下高效利用带宽；改进BBR能够显著提高长RTT流的带宽探测频率。

关键词: 拥塞控制算法, BBR优化, 低延迟, BBR公平性, 带宽探测

Abstract: The traditional congestion control algorithm based on packet loss can’t meet the requirements of many applications for network performance because of its high packet loss rate and buffer expansion. The BBR (bottleneck bandwidth and round trip) algorithm proposed by Google has attracted extensive attention and research because of its characteristics of anti packet loss, high bandwidth utilization and low delay. However, BBR still has some problems, such as high queuing delay, poor performance in a small RTT (round trip time) environment, untimely bandwidth detection, etc. This paper analyzes the queuing delay and convergence of BBR, and then puts forward an improved method: Limit inflight data, and reduce the congestion window size timely according to the network feedback to reduce the delay; In small RTT environment, the bandwidth estimation before the probe RTT stage is retained to after probe RTT; Set the maximum holding time of steady state, exit the steady cycle in time and enter the detection cycle. The simulation results in NS3 show that the improved BBR reduces the RTT and its jitter, and improves the convergence speed of the algorithm; The bandwidth can be efficiently used in the environment with small RTT; The improved BBR can significantly improve the bandwidth probe frequency of long RTT streams.

Key words: congestion control algorithm, BBR optimization, low latency, BBR fairness, bandwidth probe

黄宏平, 朱小勇, 王志远, . BBR拥塞控制算法延迟及带宽探测优化[J]. 计算机与现代化, 2022, 0(10): 113-120.

HUANG Hong-ping, ZHU Xiao-yong, WANG Zhi-yuan, . Improving Latency and Bandwidth Probe of BBR Congestion Control Algorithm[J]. Computer and Modernization, 2022, 0(10): 113-120.

参考文献［27］

［1］	JACOBSON V. Congestion avoidance and control［J］. ACM SIGCOMM Computer Communication Review, 1988,18(4):314-329.
［2］	WINSTEIN K, BALAKRISHNAN H. TCP ex machina: Computer-generated congestion control［J］. ACM SIGCOMM Computer Communication Review, 2013,43(4):123-134.
［3］	DONG M, LI Q X, ZARCHY D, et al. PCC: Re-architecting congestion control for consistent high performance［C］// Proceedings of the 12th USENIX Symposium on Networked Systems Design and Implementation(NSDI 15). 2015:395-408.
［4］	ARUN V, BALAKRISHNAN H. Copa: Practical delay-based congestion control for the Internet［C］// Proceedings of the 15th USENIX Symposium on Networked Systems Design and Implementation (NSDI 18). 2018:329-342.
［5］	CARDWELL N, CHENG Y C, GUNN C S, et al. BBR: Congestion-based congestion control: Measuring bottleneck bandwidth and round-trip propagation time［J］. Queue, 2016,14(5):20-53.
［6］	CAO Y, JAIN A, SHARMA K, et al. When to use and when not to use BBR: An empirical analysis and evaluation study［C］// Proceedings of the 2019 Internet Measurement Conference. 2019:130-136.
［7］	KLEINROCK L. Power and deterministic rules of thumb for probabilistic problems in computer communications［C］// Proceedings of the 1979 International Conference on Communications. 1979,3:1-43.
［8］	GETTYS J. Bufferbloat: Dark buffers in the Internet［J］. IEEE Internet Computing, 2011,15(3):96.
［9］	JAFFE J M. Flow control power is nondecentralizable［J］. IEEE Transactions on Communications, 1981,29(9):1301-1306.
［10］	ELBAMBY M S, PERFECTO C, BENNIS M, et al. Towards low-latency and ultra-reliable virtual reality［J］. IEEE Network, 2018,32(2):78-84.
［11］	LYNN T, GHOSAL D. TCP D*: A low latency first congestion control algorithm［J］. arXiv preprint arXiv:2012.14996, 2020.
［12］	HOCK M, BLESS R, ZITTERBART M. Experimental evaluation of BBR congestion control［C］// Proceedings of the 2017 IEEE 25th International Conference on Network Protocols (ICNP). 2017. DOI: 10.1109/ICNP.2017.8117540.
［13］	CARDWELL N. BBR bandwidth based convergence［EB/OL］. (2019-12-01)［2021-11-02］. https://github.com/google/bbr/tree/master/Documentation.
［14］	TAO Y C, JIANG J J, MA S Y, et al. Unraveling the RTT-fairness problem for BBR: A queueing model［C］// Proceedings of the 2018 IEEE Global Communications Conference (GLOBECOM). 2018. DOI: 10.1109/GLOCOM.2018.8647260.
［15］	YANG M, YANG P, WEN C Z, et al. Adaptive-BBR: Fine-grained congestion control with improved fairness and low latency［C］// Proceedings of the 2019 IEEE Wireless Communications and Networking Conference (WCNC). 2019. DOI: 10.1109/WCNC.2019.8885527.
［16］	PAN W S, TAN H B, LI X R, et al. Improved RTT fairness of BBR congestion control algorithm based on adaptive congestion window［J］. Electronics, 2021,10(5). DOI: 10.3390/electronics10050615.
［17］	ZHANG M L, MEZZAVILLA M, RANGAN S, et al. Improving Google’s BBR for reduced latency and increased fairness［C］// Proceedings of the 2019 IEEE Conference on Computer Communications Workshops (INFOCOM WKSHPS). 2019.
［18］	SONG Y J, KIM G H, CHO Y Z. BBR-CWS: Improving the inter-protocol fairness of BBR［J］. Electronics, 2020,9(5). DOI: 10.3390/electronics9050862.
［19］	CARDWELL N, CHENG Y C, YEGANEH S H, et al. BBRv2: A model-based congestion control［C］// Proceedings of the 2019 Presentation in ICCRG at IETF 104th Meeting. 2019.
［20］	SONG Y J, KIM G H, MAHMUD I, et al. Understanding of BBRv2: Evaluation and comparison with BBRv1 congestion control algorithm［J］. IEEE Access, 2021,9:37131-37145.
［21］	WANG Z Y, NI H, ZHU X Y, et al. Enhanced slow-start (ESS): Improving bandwidth utilization by estimating bandwidth delay product［J］. International Journal of Innovative Computing, Information and Control, 2021,17(6):1907-1920.
［22］	CARDWELL N. Linux TCP BBR patches for higher WiFi throughput and lower queuing delays［EB/OL］. (2018-04-05)［2021-11-15］. https://groups.google.com/forum/#!topic/bbr-dev/8pgyOyUavvY.
［23］	CHIARIOTTI F, ZANELLA A, KUCERA S, et al. BBR-S: A low-latency BBR modification for fast-varying connections［J］. IEEE Access, 2021,9:76364-76378.
［24］	DOG250-ANONYMOUS. 漫谈TCP BBR的收敛动力学(convergence dynamics)［EB/OL］. (2021-03-13)［2021-11-15］. https://zhuanlan.zhihu.com/p/356776520.
［25］	董瀚泽,郭志川. BBR拥塞控制算法在无线网络中的性能改进［J］. 哈尔滨工业大学学报, 2019,51(11):63-67.
［26］	宋雅琴,孙鹏,朱小勇. ICN拥塞控制机制研究综述［J］. 网络新媒体技术, 2021,10(5):1-8.
［27］	赵伟丰. 基于RTT的端到端网络拥塞控制研究［D］. 天津:天津大学, 2013.