面向流量预测的时间相关图卷积网络构建方法

doi:10.19665/j.issn1001-2400.20221103

摘要/Abstract

摘要：

现有数据中心虚拟网络中流量预测方法难以表征链路之间相关性,导致数据中心网络流量预测精度难以提升。基于此,提出了一种时间相关图卷积神经网络(TC-GCN),使能数据中心网络链路流量的时间和空间相关性表征,提升了流量预测精度。首先,构建具有时间属性的图卷积神经网络邻接矩阵,解决虚拟网络链路间流量异步性导致的预测偏差问题,实现了链路相关性的精准表征;其次,设计基于长/短窗口图卷积神经网络加权的流量预测机制,利用有限长度长/短窗口适配流量序列的平滑段与波动段,有效避免了神经网络梯度消失问题,提升了虚拟网络的流量预测精度;最后,设计了一个误差加权单元对长短窗口图卷积神经网络的预测结果进行加权求和,该网络的输出即为链路流量的预测值。为保障结果的实用性,基于真实的数据中心网络数据对所提时间相关图卷积网络进行了仿真实验。实验结果表明,所提预测方法相比于传统的图卷积神经网络流量预测方法具有更高的预测精度。

关键词: 虚拟化技术, 网络拓扑结构, 图卷积神经网络, 流量预测

Abstract:

The existing traffic prediction methods in the virtual network of data centers characterize the correlation between links with difficulty,which leads to the difficulty in improving the accuracy of traffic prediction.Based on this,this paper proposes a Temporal Correlation Graph Convolutional neural Network (TC-GCN),which enables the representation of Temporal and spatial Correlation of the data center Network link traffic and improves the accuracy of traffic prediction.First,the graph convolutional neural network adjacency matrix with the time attribute is constructed to solve the problem of prediction deviation caused by traffic asynchronism between virtual network links,and to achieve accurate representation of link correlation.Second,a traffic prediction mechanism based on long/short window graph convolutional neural network weighting is designed,which adapts the smooth and fluctuating segments of the traffic sequence with a finite length long/short window,effectively avoids the vanishing gradient problem of the neural network,and improves the traffic prediction accuracy of the virtual network.Finally,an error weighting unit is designed to sum the prediction results of the long/short window graph convolutional neural network.The output of the network is the predicted value of link traffic.In order to ensure the practicability of the results,the simulation experiments of the proposed temporal correlation graph convolutional network are carried out based on the real data center network data.Experimental results show that the proposed method has a higher prediction accuracy than the traditional graph convolutional neural network traffic prediction method.

Key words: virtualization, network topology, graph convolutional network, traffic prediction

中图分类号:

TN915.03

张可涵,李红艳,刘文慧,王鹏. 面向流量预测的时间相关图卷积网络构建方法[J]. 西安电子科技大学学报, 2023, 50(5): 11-20.

ZHANG Kehan,LI Hongyan,LIU Wenhui,WANG Peng. Construction method of temporal correlation graph convolution network for traffic prediction[J]. Journal of Xidian University, 2023, 50(5): 11-20.

图/表 8

图1

图2

图3

图4

图5

图6

表1

图7

参考文献 28

[1]	王春晖. “东数西算”工程的战略布局[J]. 中国电信业, 2021(8):47-51.
	WANG Chunhui. Strategic Layout of "East Digital West Computing" Project[J]. China Telecom Industry, 2021(8):47-51.
[2]	LIU Y, MUPPALA J K, VEERARAGHAVAN M, et al. Data Center Networks[M]. Heidelberg: Springer International Publishing, 2013.
[3]	王绪. 面向数据密集型计算的高效能可重构计算系统结构研究与设计[D]. 上海: 上海交通大学, 2020.
[4]	JANARDHANAN D, BARRETT E. CPU Workload Forecasting of Machines in Data Centers Using LSTM Recurrent Neural Networks and ARIMA Models[C]//2017 12th International Conference for Internet Technology and Secured Transactions (ICITST).Piscataway:IEEE, 2017:55-60.
[5]	SHIRZADI M. Representing Increasing Virtual Machine Security Strategy in Cloud Computing Computations[J]. Electrical Science & Engineering, 2021, 3(2):7-16.
[6]	XING Y, ZHAN Y. Virtualization and Cloud Computing[M]. Heidelberg:Springer, 2012:305-312.
[7]	CHOWDHURY N MM K, BOUTABA R. A Survey of Network Virtualization[J]. Computer Networks, 2010, 54(5): 862-876. doi: 10.1016/j.comnet.2009.10.017
[8]	HAIDER A, POTTER R, NAKAO A, et al. Challenges in Resource Allocation in Network Virtualization[C]//20th ITC Specialist Seminar. Ghent:ITC, 2009:1-9.
[9]	马悦, 张玉梅. 面向多接入边缘计算的VNFM分布式部署方案[J]. 西安电子科技大学学报, 2021, 48(4):20-26.
	MA Yue, ZHANG Yumei. Method for Distributed Deployment of the Virtual Network Function Manager for MEC[J]. Journal of Xidian University, 2021, 48(4):20-26.
[10]	AL-FARES M, LOUKISSAS A, VAHDAT A, et al. A Scalable,Commodity Data Center Network Architecture[J]. ACM SIGCOMM Computer Communication Review, 2008, 38(4):63-74.
[11]	VERBRAEKEN J, WOLTING M, KATZY J, et al. A Survey on Distributed Machine Learning[J]. ACM Computing Surveys (CSUR), 2020, 53(2):1-33.
[12]	YU A, YANG H, BAI W, et al. LeveragingDeep Learning to Achieve Efficient Resource Allocation with Traffic Evaluation in Datacenter Optical Networks[C]//2018 Optical Fiber Communications Conference and Exposition (OFC).Piscataway:IEEE, 2018:1-3.
[13]	YU A, YANG H, YAO Q, et al. Scheduling withFlow Prediction Based on Time and Frequency 2D Classification for Hybrid Electrical/Optical Intra-Datacenter Networks[C]//2019 Optical Fiber Communication Conference and Exhibition.Piscataway:IEEE, 2019:1-3.
[14]	张杰钧. 数据中心资源分配机制研究[D]. 北京: 北京邮电大学, 2019.
[15]	TANG L, CHEN H. Joint Pricing and Capacity Planning in the IaaS Cloud Market[J]. IEEE Transactions on Cloud Computing, 2014, 5(1):57-70. doi: 10.1109/TCC.2014.2372811
[16]	JIANG C, QIU Y, SHI W, et al. Characterizing Co-Located Workloads in Alibaba Cloud Datacenters[J]. IEEE Transactions on Cloud Computing, 2020, PP(99):1-1.
[17]	KO T, RAZA S M, DANG T B, et al. Network Prediction with Traffic Gradient Classification Using Convolutional Neural Networks[C]//2020 14th International Conference on Ubiquitous Information Management and Communication (IMCOM).Piscataway:IEEE, 2020:1-4.
[18]	许缓缓, 李洪梅, 李富余, 等. 一种时空卷积的步态识别方法[J]. 西安电子科技大学学报, 2021, 48(4):144-150.
	XU Huanhuan, LI Hongmei, LI Fuyu, et al. Gait Recognition Method Based on Spatial-Temporal Convolution[J]. Journal of Xidian University, 2021, 48(4):144-150.
[19]	ZAREMBA W, SUTSKEVER I, VINYALS O, et al. Recurrent Neural Network Regularization (2014)[J/OL].[2014-09-08]. https://arxiv.org/abs/1409.2329v1.
[20]	LIU L, QIU Z, LI G, et al. Contextualized Spatial-Temporal Network for Taxi Origin-Destination Demand Prediction[J]. IEEE Transactions on Intelligent Transportation Systems, 2019, 20(10):3875-3887. doi: 10.1109/TITS.6979
[21]	JIE Z A, GC A, SH A, et al. Graph Neural Networks:A Review of Methods and Applications[J]. AI Open, 2020, 1:57-81. doi: 10.1016/j.aiopen.2021.01.001
[22]	VELIKOVI P, CUCURULL G, CASANOVA A, et al. Graph Attention Networks[C]//International Conference on Learning Representations. La Jolla: ICLR, 2018:1-12.
[23]	DEFFERRARD M, BRESSON X, VANDERGHEYNST P, et al. Convolutional Neural Networks on Graphs with Fast Localized Spectral Filtering[J]. Advances in Neural Information Processing Systems, 2016, 29:1-9.
[24]	KIPF T N, WELLING M. Semi-Supervised CLassification with Graph Convolutional Networks (2016)[J/OL].[2016-09-09]. https://arxiv.org/abs/1609.02907.
[25]	ZHAO L, SONG Y, ZHANG C, et al. T-GCN:A Temporal Graph Convolutional Network for Traffic Prediction[J]. IEEE Transactions on Intelligent Transportation Systems, 2019, PP(99):1-11.
[26]	CHO K, MERRIENBOER B V, GULCEHRE C, et al. Learning Phrase Representations Using RNN Encoder-Decoder for Statistical Machine Translation[J]. Computer Science, 2014(4):1-15.
[27]	薛可, 李增智, 刘浏, 等. 基于ARIMA模型的网络流量预测[J]. 微电子学与计算机, 2004(7):84-87.
	XUE Ke, LIZengzhi, Liu Liu, et al. Song Chengqian.Network Traffic Prediction Based on ARIMA Model[J]. Microelectronics & Computer, 2004(7):84-87.
[28]	GUO S, LIN Y, FENG N, et al. AttentionBased Spatial-Temporal Graph Convolutional Networks for Traffic Flow Forecasting[C]//Proceedings of the AAAI Conference on Artificial Intelligence. Palo Alto: AAAI, 2019:922-929.

模型	R_Accuracy	R_MAE	R_RMSE
ARIMA	0.619 3	15.032 0	17.281 2
GCN	0.624 1	14.925 7	16.802 7
LSTM	0.728 4	13.873 3	15.205 2
STGCN	0.800 7	9.590 3	11.002 6
T-GCN	0.859 4	8.409 7	10.576 6
TC-GCN	0.906 0	7.547 6	8.840 0