MgdFlow：微电网场景下的多粒度数据流管理算法

doi:10.11959/j.issn.1000-436x.2023175

摘要/Abstract

摘要：

针对软件定义网络（SDN）设备被动响应式插入流表项导致表项空间占用和资源调度频繁的问题，提出了一种基于业务功能的多粒度数据流管理算法（MgdFlow）。通过延迟排序和多重聚合与分流，减少了流表项插入的数目和调度指令的次数。实验数据表明，所提算法在负载均衡性能方面比自适应两级流表提升18%，在平均表项占用方面比传统OpenFlow方案提升9%，在定义的控制器调度资源比率ImproveQoS方面比2种对比方案提升24%和12%。

关键词: 软件定义网络, 流表管理, 负载均衡, 服务质量

Abstract:

Aiming at the problem of frequent entry space occupation and resource scheduling caused by passive response insertion of flow entries by software defined network (SDN) devices, a operational function-based multi-granularity data flow management algorithm (MgdFlow) was proposed.Through delayed sorting and multiple aggregation and splitting, the number of flow entry insertions and scheduling instructions were reduced.Experimental data shows that the load balancing performance is 18% higher than the adaptive two-stage flow, and the average entry occupation is 9% lower than the traditional OpenFlow scheme.There are also 24% and 12% improvements in the defined controller scheduling resource ratio ImproveQoS compared to other schemes.

Key words: software defined network, flow table management, load balancing, QoS

中图分类号:

TP393.02

荆有波, 曹清越, 朱瑞. MgdFlow：微电网场景下的多粒度数据流管理算法[J]. 通信学报, 2023, 44(10): 94-102.

Youbo JING, Qingyue CAO, Rui ZHU. MgdFlow: multi-granularity data flow management algorithm in microgrid scenario[J]. Journal on Communications, 2023, 44(10): 94-102.

图/表 10

图1

图2

表1

MgdFlow模型的符号及含义"

符号	含义
s_h	第h台交换机
l_q	交换机之间的第q条链路
Rate _f	流f的速率
fs_n	对网络中的流聚合成的第n个流集
route_m	网络设备和链路组成的第m条路由
so _j	粗粒度网络中第j台交换机的表项
so_k	细粒度网络中第k台交换机的表项
$a_{r o u t e}^{f s}$	若流集fs_n要通过路由Route，则为1，否则为0
$b_{r o u t e_{m^{'}}^{f s'}}^{f s}$	若流集fs＇_g 要经过路由Route＇，则为1，否则为0

表1

表2

表3

图3

图4

图5

图6

表4

参考文献 27

[1]	郭金明, 李欣然, 邓威 ,等. 基于2 层规划的间歇性分布式电源及无功补偿综合优化配置[J]. 中国电机工程学报, 2013,33(28): 25-33,6.
	GUO J M , LI X R , DENG W ,et al. Comprehensive optimal allocation of intermittent distributed generation and reactive power compensation based on bilevel planning[J]. Proceedings of the CSEE, 2013,33(28): 25-33,6.
[2]	陈其森, 汪湘晋, 池伟 ,等. 多微电网互联系统能量管理方法研究[J]. 电力系统保护与控制, 2018,46(11): 83-91.
	CHEN Q S , WANG X J , CHI W ,et al. Research on energy management method of multi-microgrids[J]. Power System Protection and Control, 2018,46(11): 83-91.
[3]	曹正斐, 张忠辉, 董治成 ,等. 基于区块链的多互联微电网分布式协调优化调度[J]. 电力系统及其自动化学报, 2022,34(9): 138-145.
	CAO Z F , ZHANG Z H , DONG Z C ,et al. Distributed coordination and optimization scheduling of multi-interconnected microgrids based on block chain[J]. Proceedings of the CSU-EPSA, 2022,34(9): 138-145.
[4]	郭泽华, 窦松石, 齐力 ,等. 面向软件定义广域网的路径可编程性保障研究综述[J]. 电子与信息学报, 2023,45(5): 1899-1910.
	GUO Z H , DOU S S , QI L ,et al. A survey of maintaining the path programmability in software-defined wide area networks[J]. Journal of Electronics ＆ Information Technology, 2023,45(5): 1899-1910.
[5]	张朝昆, 崔勇, 唐翯祎 ,等. 软件定义网络(SDN)研究进展[J]. 软件学报, 2015,26(1): 62-81.
	ZHANG C K , CUI Y , TANG H Y ,et al. State-of-the-art survey on software-defined networking (SDN)[J]. Journal of Software, 2015,26(1): 62-81.
[6]	左青云, 陈鸣, 赵广松 ,等. 基于 OpenFlow 的 SDN 技术研究[J]. 软件学报, 2013,24(5): 1078-1097.
	ZUO Q Y , CHEN M , ZHAO G S ,et al. Research on OpenFlow-based SDN technologies[J]. Journal of Software, 2013,24(5): 1078-1097.
[7]	张进进 . 基于SDN的电力通信网QoS技术研究[D]. 重庆:重庆邮电大学, 2020.
	ZHANG J J . Research on QoS technology of power communication network based on SDN[D]. Chongqing:Chongqing University of Posts and Telecommunications, 2020.
[8]	陈志鹏, 徐明伟, 杨芫 . SDN交换机转发规则TCAM存储优化综述[J]. 计算机学报, 2021,44(7): 1341-1362.
	CHEN Z P , XU M W , YANG Y . A survey on TCAM storage optimization for SDN switch forwarding rules[J]. Chinese Journal of Computers, 2021,44(7): 1341-1362.
[9]	谢升旭, 邢长友, 张国敏 ,等. OpenFlow交换机流表溢出缓解技术研究综述[J]. 计算机研究与发展, 2021,58(7): 1544-1562.
	XIE S X , XING C Y , ZHANG G M ,et al. Survey of OpenFlow switch flow table overflow mitigation techniques[J]. Journal of Computer Research and Development, 2021,58(7): 1544-1562.
[10]	SHIRALI-SHAHREZA S , GANJALI Y . Delayed installation and expedited eviction:an alternative approach to reduce flow table occupancy in SDN switches[J]. IEEE/ACM Transactions on Networking, 2018,26(4): 1547-1561.
[11]	GUO Z H , XU Y , LIU Y F ,et al. AggreFlow:achieving power efficiency,load balancing,and quality of service in data center networks[J]. IEEE/ACM Transactions on Networking, 2021,29(1): 17-33.
[12]	SOLIMAN M , NANDY B , LAMBADARIS I ,et al. Exploring source routed forwarding in SDN-based WANs[C]// Proceedings of 2014 IEEE International Conference on Communications (ICC). Piscataway:IEEE Press, 2014: 3070-3075.
[13]	GUO Z H , XU Y , CELLO M ,et al. JumpFlow:reducing flow table usage in software-defined networks[J]. Computer Networks, 2015,92: 300-315.
[14]	李佳 . 面向 SDN 的流表超时和路由算法研究[D]. 广州:华南理工大学, 2020.
	LI J . Research on flow table timeout and routing algorithm for SDN[D]. Guangzhou:South China University of Technology, 2020.
[15]	刘振鹏, 李明, 王鑫鹏 ,等. 基于时序与集合的SDN流表更新策略[J]. 河北大学学报(自然科学版), 2020,40(4): 427-432.
	LIU Z P , LI M , WANG X P ,et al. Update strategy of SDN flow table based on time series and sets[J]. Journal of Hebei University (Natural Science Edition), 2020,40(4): 427-432.
[16]	曾友雯, 李双庆, 邹东升 . 采用 OpenFlow 交换机的服务器负载均衡策略[J]. 重庆大学学报, 2021,44(11): 48-56.
	ZENG Y W , LI S Q , ZOU D S . Server load balancing strategy using OpenFlow switch[J]. Journal of Chongqing University (Natural Science Edition), 2021,44(11): 48-56.
[17]	马晓航 . 基于流表项超时的 SDN 网络性能优化[D]. 桂林:桂林电子科技大学, 2021.
	MA X H . Performance optimization of SDN network based on timeout of flow table entry[D]. Guilin:Guilin University of Electronic Technology, 2021.
[18]	严可意 . 基于机器学习的 SDN 流表优化研究与实现[D]. 北京:北京邮电大学, 2021.
	YAN K Y . Research and implementation of optimization of SDN flow table based on machine learning[D]. Beijing:Beijing University of Posts and Telecommunications, 2021.
[19]	齐婵, 刘建伟, 毛剑 ,等. 基于分类和时序的SDN流表更新一致性方案[J]. 计算机应用研究, 2018,35(11): 3405-3408,3412.
	QI C , LIU J W , MAO J ,et al. Classification and sequence based consistent flow update scheme in SDN[J]. Application Research of Computers, 2018,35(11): 3405-3408,3412.
[20]	PETALE S , THANGARAJ J . Failure-based controller placement in software defined networks[J]. IEEE Transactions on Network and Service Management, 2020,17(1): 503-516.
[21]	KAMATH S , SRIVASTAVA A , KAMATH P ,et al. Application aware multiple constraint optimal paths for transport network using SDN[J]. IEEE Transactions on Network and Service Management, 2021,18(4): 4376-4390.
[22]	CHANG Y C , LIN H T , CHU H M ,et al. Efficient topology discovery for software-defined networks[J]. IEEE Transactions on Network and Service Management, 2021,18(2): 1375-1388.
[23]	LAI W K , WANG Y C , CHEN Y C ,et al. TSSM:time-sharing switch migration to balance loads of distributed SDN controllers[J]. IEEE Transactions on Network and Service Management, 2022,19(2): 1585-1597.
[24]	HAMID G , BARBETTE T , KATSIKA G P ,et al. Packet order matters! improving application performance by deliberately delaying packets[C]// Proceedings of the 19th USENIX Symposium on Networked Systems Design and Implementation. Berkeley:USENIX Association, 2022: 807-827.
[25]	ISYAKU B , ABU B K B , GHALEB F A ,et al. Performance evaluation of flowtable eviction mechanisms for software defined networks considering traffic flows variabilities[C]// Proceedings of 2022 IEEE 12th Symposium on Computer Applications ＆ Industrial Electronics (ISCAIE). Piscataway:IEEE Press, 2022: 71-75.
[26]	QIU K , YUAN J , ZHAO J ,et al. FastRule:efficient flow entry updates for TCAM-based OpenFlow switches[J]. IEEE Journal on Selected Areas in Communications, 2019,37(3): 484-498.
[27]	SHANG Z H , WU H , WOLTER K . Buffer management for reducing packet-in messages in OpenFlow networks[C]// Proceedings of 2019 IEEE 11th International Conference on Communication Software and Networks (ICCSN). Piscataway:IEEE Press, 2019: 458-465.

软件名称	版本号
Mininet	2.3.0
OpenDayLight	15.0.0
Open vSwitch	2.10.0
Apache Maven	3.6.3

硬件名称	型号
主机CPU	Intel(R) Core(TM)i5-7200U 2.5 GHz
主机内存	16GB DDR4 2133 MHz
主机系统	Windows10 专业版
虚拟机内存	4 GB
虚拟机硬盘	100 GB
虚拟机系统	Ubuntu16.04

方案	负载均衡	平均流表项占有率	ImproveQoS
传统OpenFlow方案	53%	36%	47%
自适应两级流表方案	46%	43%	35%
MgdFlow	28%	27%	23%