大型IP网络流量矩阵分析预测的探讨研究

doi:10.12305/j.issn.1001-506X.2024.06.35

摘要/Abstract

摘要：

高效、准确的网际协议(internet protocol, IP)网络流量流向分析预测是网络规划建设的基础。通过部署流量采集分析系统, 运营商可轻松获取网络总流量、节点流量、节点分方向流量等较完备的历史基础数据, 为流量分析预测提供关键的输入。IP网络流量分析预测方法主要包括两类: 传统统计模型和神经网络模型, 近年提出的NeuralProphet模型因结合两者优点而得到广泛关注和应用。首次基于NeuralProphet模型对大型运营级IP网络源节点到目的节点的流量流向进行直接预测, 并采用改进的损失函数优化模型训练, 预测结果表明NeuralProphet模型能够更科学、准确地预测IP网络流量矩阵, 整体预测精度提升了8.7%, 同时模型扩展性和鲁棒性也具有更佳的表现, 可以更好地满足IP网络规划建设和运行维护的实际需求。

关键词: 流量矩阵, 源节点到目的节点流量流向, 节点流量, 预测模型, 自回归

Abstract:

Efficient and accurate analysis and prediction of traffic flow direction for Internet protocol (IP) network are the basis of network planning and construction. By deploying a traffic collection and analysis system, operators can easily obtain comprehensive historical data such as network total traffic, node traffic, and node directional traffic, which provides key inputs for traffic analysis and prediction. Methods of traffic analysis and prediction for IP network are generally divided into two categories: traditional statistical model and neural network model. The NeuralProphet model proposed in recent years has been widely applied due to its combination of the advantages of the above models. It is the first time to directly predict the origin-destination traffic flow of large-scale carrier-grade IP network based on the NeuralProphet model, and adopts the improved loss function to optimize model training. The prediction results show that the NeuralProphet model can predict traffic matrix of IP network more scientifically and accurately, and the overall prediction accuracy was improved by 8.7%. Meanwhile, the model has better scalability and robustness, which can better meet the actual needs of IP network planning and maintenance.

Key words: traffic matrix, origin-destination traffic flow, node traffic, prediction model, auto-regression

中图分类号:

TN915

韦烜, 刘志华, 李青, 何晓明, 黄君雅. 大型IP网络流量矩阵分析预测的探讨研究[J]. 系统工程与电子技术, 2024, 46(6): 2164-2173.

Xuan WEI, Zhihua LIU, Qing LI, Xiaoming HE, Junya HUANG. Research on analysis and prediction of traffic matrix for large-scale IP network[J]. Systems Engineering and Electronics, 2024, 46(6): 2164-2173.

图/表 14

图1

表1

表2

图2

图3

图4

图5

图6

图7

图8

图9

图10

图11

表3

参考文献 33

1	SAFRIANTI E, SARI L O, SARI N A. Real-time network device monitoring system with simple network management protocol (SNMP) Model[C]//Proc. of the 3rd International Confe-rence on Research and Academic Community Services, 2021: 122-127.
2	SHMELKIN I, SPRINGER T. On adapting SNMP as communication protocol in distributed control loops for self-adaptive systems[C]//Proc. of the IEEE International Conference on Autonomic Computing and Self-organizing Systems, 2021: 61-70.
3	董兴强, 李晓冰. 电信运营商网络流量采集模型研究及应用[J]. 移动通信, 2020, 44 (3): 67- 71.
	DONG X Q , LI X B . Research and application of telecommunication operator network traffic collection model[J]. Mobile Communications, 2020, 44 (3): 67- 71.
4	YANG B W, LIU D. Design of IP network traffic acquisition system based on xFlow[C]//Proc. of the IEEE 8th Joint International Information Technology and Artificial Intelligence Conference, 2019: 1631-1634.
5	NIE L S , WANG H Z , JIANG X , et al. Traffic measurement optimization based on reinforcement learning in large-scale ITS-oriented backbone networks[J]. IEEE Access, 2020, 8, 36988- 36996. doi: 10.1109/ACCESS.2020.2975238
6	WANG M , LU Y Q , QIN J C . Source-based defense against DDoS attacks in SDN based on sFlow and SOM[J]. IEEE Access, 2021, 10, 2097- 2116.
7	荣红佳, 盛虎, 闫秋婷. 基于改进R/S估计算法的网络流量长相关性分析[J]. 大连交通大学学报, 2021, 42 (2): 114- 119.
	RONG H J , SHENG H , YAN Q T . Analysis of network traffic long correlation based on improved R/S estimation algorithm[J]. Journal of Dalian Jiaotong University, 2021, 42 (2): 114- 119.
8	ERRAMILLI A , ROUGHAN M , VEITCH D , et al. Self-similar traffic and network dynamics[J]. Proceedings of the IEEE, 2002, 90 (5): 800- 819. doi: 10.1109/JPROC.2002.1015008
9	OLGA V , FERNANDO R , LUIS J H , et al. New developments in time series and forecasting[J]. Engineering Proceedings, 2023, 39 (1): 135- 148.
10	王婧, 鲍贵. 贝叶斯统计与传统统计方法的比较[J]. 统计与决策, 2021, 37 (1): 24- 29.
	WANG J , BAO G . Comparison between bayesian statistics and traditional statistical methods[J]. Statistics & Decision, 2021, 37 (1): 24- 29.
11	LIU H , ZHANG X Y , YANG Y X , et al. Hourly traffic flow forecasting using a new hybrid modelling method[J]. Journal of Central South University, 2022, 29 (4): 1389- 1402. doi: 10.1007/s11771-022-5000-2
12	LIAO L C , HU Z Y , HSU C Y , et al. Fourier graph convolution network for time series prediction[J]. Mathematics, 2023, 11 (7): 122- 131.
13	XU G Q , XIA C S , QIAN J , et al. A network traffic prediction algorithm based on Prophet-EALSTM-GPR[J]. Journal on Internet of Things, 2023, 4 (2): 173- 182.
14	VACCARI I , CARLEVARO A , NARTENI S , et al. Xplainable and reliable against adversarial machine learning in data analytics[J]. IEEE Access, 2022, 10, 83949- 83970.
15	史朝卫, 孟相如, 康巧燕, 等. 基于混合流量预测的虚拟网络拓扑重构方法[J]. 系统工程与电子技术, 2021, 43 (5): 1382- 1388.
	SHI C W , MENG X R , KANG Q Y , et al. Virtual network topology reconfiguration approach based on hybrid traffic[J]. Systems Engineering and Electronics, 2021, 43 (5): 1382- 1388.
16	YAO E Z , ZHANG L J , LI X H , et al. Traffic forecasting of back servers based on ARIMA-LSTM-CF hybrid model[J]. International Journal of Computational Intelligence Systems, 2023, 16 (1): 244- 256.
17	王菁, 文晓东, 王春枝. 基于动态扩散卷积交互图神经网络的网络流量预测[J]. 计算机应用研究, 2023, 40 (1): 97- 101.
	WANG J , WEN X D , WANG C Z . Network traffic prediction based on dynamic diffusion convolutional interaction graph neural network[J]. Application Research of Computers, 2023, 40 (1): 97- 101.
18	YANG Y G , GENG S P , ZHANG B C , et al. Long term 5G network traffic forecasting via modeling non-stationarity with deep learning[J]. Communications Engineering, 2023, 2 (1): 135- 142.
19	ETNGU R , TAN C S , CHEE T C , et al. AI-assisted traffic matrix prediction using GA-enabled deep ensemble learning for hybrid SDN[J]. Computer Communications, 2023, 203 (2): 1124- 1131.
20	RAU F , SOTO I , ZABALABLANCO D , et al. A novel traffic prediction method using machine learning for energy efficiency in service provider networks[J]. Sensors, 2023, 23 (11): 155- 167.
21	SWETHA K, PRABU U, ANGEL G, et al. A study on traffic matrix estimation techniques in software-defined networks[C]//Proc. of the 6th International Conference on Electronics, Communication and Aerospace Technology, 2022: 604-611.
22	JIANG D D, HU G M. A novel approach to large-scale IP traffic matrix estimation based on RBF neural network[C]//Proc. of the 4th International Conference on Wireless Communications, Networking and Mobile Computing, 2008.
23	TRIEBE O, HEWAMALAGE H, PILYUGINA P, et al. NeuralProphet: explainable forecasting at scale[EB/OL]. [2023-04-10]. https://arxiv.org/pdf/2111.15397.pdf.
24	TAYLOR S J , LETHAM B . Forecasting at scale[J]. The American Statistician, 2018, 72 (1): 37- 45.
25	孙慧慧, 刘强. 基于改进Huber损失的部分线性模型稳健经验似然推断[J]. 系统科学与数学, 2022, 42 (5): 1330- 1343.
	SUN H H , LIU Q . Robust empirical likelihood inference of partial linear models based on improved huber loss[J]. Systems Science and Mathematical Sciences, 2022, 42 (5): 1330- 1343.
26	WU X Y , FU S D , HE Z J . Research on short-term traffic flow combination prediction based on CEEMDAN and machine learning[J]. Applied Sciences, 2022, 13 (1): 1121- 1132.
27	JEBA N , RATHI S . Attention-based multiscale spatiotemporal network for traffic forecast with fusion of external factors[J]. ISPRS International Journal of Geo-Information, 2022, 11 (12): 323- 330.
28	LIN L , LI W Z , ZHU L . Data-driven graph filter-based graph convolutional neural network approach for network-level multi-step traffic prediction[J]. Sustainability, 2022, 14 (24): 211- 219.
29	DALAL A , IMTIAZ A , EBRAHIM A . Deep learning based network traffic matrix prediction[J]. International Journal of Intelligent Networks, 2021, 2 (1): 135- 142.
30	YANG W C , RUI H , ZHAO Q H . A sequence-to-sequence traffic predictor on software-defined networking[J]. International Journal of Web and Grid Services, 2021, 17 (3): 1210- 1221.
31	SAYED S A , YASSER H A , AHMED H H . Artificial intelligence-based traffic flow prediction: a comprehensive review[J]. Journal of Electrical Systems and Information Technology, 2023, 10 (1): 144- 153.
32	XIONG P P , CHEN S T , YAN S L . Time-delay nonlinear model based on interval grey number and its application[J]. Journal of Systems Engineering and Electronics, 2022, 33 (2): 370- 380.
33	LYU S T , LI X H , FAN T , et al. Deep learning for fast channel estimation in millimeter-wave MIMO systems[J]. Journal of Systems Engineering and Electronics, 2022, 33 (6): 1088- 1095.

节点	节点
节点	N1	N2	N3	…	N14	N15	N16
N1	437.32	28.08	14.98	…	21.01	6.69	52.06
N2	4.49	60.76	6.86	…	6.12	6.11	16.32
N3	21.62	27.64	306.76	…	16.01	48.22	46.34
				$\ddots$			
N14	17.08	5.69	17.18	…	805.44	8.75	256.11
N15	7.81	15.78	86.89	…	10.98	206.89	23.53
N16	51.95	18.81	57.34	…	323.07	28.82	2 337.57

时间	节点
时间	N1-N1	N1-N2	N1-N3	…	N16-N14	N16-N15	N16-N16
2015年5月	420.31	31.77	13.73	…	355.62	29.90	2 220.03
2015年6月	437.32	28.08	14.98	…	323.07	28.82	2 337.57
2015年7月	484.97	24.85	14.03	…	339.89	28.81	2 642.86
				$\ddots$			
2023年1月	2 985.36	357.42	254.48	…	504.20	93.56	3 098.55
2023年2月	2 904.61	347.41	268.89	…	511.16	94.60	3 314.75
2023年3月	2 690.06	324.63	285.16	…	594.21	99.46	3 267.70

模型	评价指标
模型	RMSE(Gb/s)	MAE(Gb/s)	MAPE/%
ARIMA	246.7	100.9	53.3
Prophet	124.6	56.9	29.2
NeuralProphet	75.1	33.1	20.5

[1]	禄晓飞, 靳硕静, 洪灵, 戴奉周. 基于聚类求解TVAR模型的目标微多普勒分析[J]. 系统工程与电子技术, 2023, 45(3): 660-668.
[2]	宋星, 贾红丽, 王谦, 赵汝东. 基于时间序列挖掘的合成旅装备维修保障能力预测[J]. 系统工程与电子技术, 2020, 42(4): 878-886.
[3]	徐丹, 符吉祥, 孙光才, 邢孟道, 苏涛. 宽带弹道高速群目标运动精补偿方法[J]. 系统工程与电子技术, 2019, 41(10): 2205-2213.
[4]	韩春耀, 熊家军, 张凯, 兰旭辉. 高超声速飞行器分解集成轨迹预测算法[J]. 系统工程与电子技术, 2018, 40(1): 151-158.
[5]	赵兴刚, 王首勇. 基于AR模型的飞机尾流距离检测方法[J]. 系统工程与电子技术, 2017, 39(4): 726-732.
[6]	刘洁瑜, 沈强, 李灿, 秦伟伟. 基于优化KF的MEMS陀螺阵列信号融合方法[J]. 系统工程与电子技术, 2016, 38(12): 2705-2710.
[7]	郑作虎，王首勇. 基于ARSαS模型参数估计的雷达目标检测方法[J]. 系统工程与电子技术, 2015, 37(4): 782-788.
[8]	赵建忠, 徐廷学, 叶文, 张磊. 基于数据融合和改进MUGM(1,m,w)的导弹装备故障预测[J]. 系统工程与电子技术, 2015, 37(4): 832-837.
[9]	张晓蓉，吴成茂，李文学. 基于多涡卷Jerk-Chua混沌和自编码的扩频码构造方法[J]. 系统工程与电子技术, 2015, 37(4): 936-941.
[10]	郭晓君，刘思峰，方志耕. 基于合成灰数灰度的区间灰数自忆性预测模型[J]. 系统工程与电子技术, 2014, 36(6): 1124-1129.
[11]	盛佳恋，张磊，邢孟道，徐刚，贺丽. 联合运动补偿的逆合成孔径雷达成像方位定标方法[J]. 系统工程与电子技术, 2014, 36(5): 859-865.
[12]	王大鹏，汪秉文，李睿凡. 考虑合成灰数灰度性质的改进区间灰数预测模型[J]. Journal of Systems Engineering and Electronics, 2013, 35(5): 1013-1017.
[13]	李星秀，韦志辉，肖亮，费选. 非局部正则化的压缩感知图像重建算法[J]. Journal of Systems Engineering and Electronics, 2013, 35(1): 196-202.
[14]	王汝夯，黄建国，张群飞. 基于分数阶傅里叶变换的LFM混响空时预白化方法[J]. Journal of Systems Engineering and Electronics, 2011, 33(7): 1523-1526.
[15]	赵琳，高帅和，丁继成. 应用ARMA模型的辅助载波多普勒精确估计[J]. Journal of Systems Engineering and Electronics, 2011, 33(5): 1085-.