Accurate and efficient elephant-flow classification based on co-trained models in evolved software-defined networks^✩

Abstract

Accurate early classification of elephant flows (elephants) is important for network management and resource optimization. Elephant models, mainly based on the byte count of flows, can always achieve high accuracy, but not in a time-efficient manner. The time efficiency becomes even worse when the flows to be classified are sampled by flow entry timeout over Software-Defined Networks (SDNs) to achieve a better resource efficiency. This paper addresses this situation by combining co-training and Reinforcement Learning (RL) to enable a closed-loop classification approach that divides the entire classification process into episodes, each involving two elephant models. One predicts elephants and is retrained by a selection of flows automatically labeled online by the other. RL is used to formulate a reward function that estimates the values of the possible actions based on the current states of both models and further adjusts the ratio of flows to be labeled in each phase. Extensive evaluation based on real traffic traces shows that the proposed approach can stably predict elephants using the packets received in the first 10% of their lifetime with an accuracy of over 80%, and using only about 10% more control channel bandwidth than the baseline over the evolved SDNs.

Keywords

Software-defined network / Flow classification / Co-training / Reinforcement learning / Flow entry timeout

References

[1] U. Deshpande, N. Rajesh, S. Dhananjaya, Survey of sdn traffic flow classification approaches, INFOCOMP J. Comput. Sci 20 (1) (2021) 7-22.

[2] R. Jang, D. Min, S. Moon, D. Mohaisen, D. Nyang, Sketchflow: per-flow systematic sampling using sketch saturation event, in: Proceeding of the IEEE INFOCOM 2020-IEEE Conference on Computer Communications, IEEE, 2020, pp. 1339-1348.

[3] F. Tang, H. Zhang, L.T. Yang, L. Chen, Elephant flow detection and differentiated scheduling with efficient sampling and classification, IEEE Trans. Cloud Comput. 9 (3) (2019) 1022-1036.

[4] O.N. Foundation, Software-defined networking: the new norm for networks, https://opennetworking.org/wp-content/uploads/2011/09/wp-sdn-newnorm.pdf, 2012. (Accessed 20 April 2023).

[5] L.X. Liao, X. Ma, Z. Li, H.-C. Chao, Dynamic flow entry timeouts based packet nonuniform sampling for elephant flow detection, in: Proceeding of the 2021 Inter-national Symposium on Computer Technology and Information Science, IEEE, 2021, pp. 431-439.

[6] I. Muhammad, Z. Yan, Supervised machine learning approaches: a survey, ICTACT J. Soft Comput. 5 (3) (2015).

[7] F. Estrada-Solano, O.M. Caicedo, N.L. Da Fonseca, Nelly: flow detection using incre-mental learning at the server side of sdn-based data centers, IEEE Trans. Ind. Inform. 16 (2) (2019) 1362-1372.

[8] R. Durner, W. Kellerer, Network function offloading through classification of ele-phant flows, IEEE Trans. Netw. Serv. Manag. 17 (2) (2020) 807-820.

[9] L.P. Kaelbling, M.L. Littman, A.W. Moore, Reinforcement learning: a survey, ICTACT J. Soft Comput. 4 (1996) 237-285.

[10] M. Baca, Introduction to Metadata third ed., Getty Publications, Los Angeles, 2016.

[11] L.X. Liao, H.-C. Chao, M.-Y. Chen, Intelligently modeling, detecting, and schedul-ing elephant flows in software defined energy cloud: a survey, J. Parallel Distrib. Comput. 146 (2020) 64-78.

[12] L. Peng, B. Yang, Y. Chen, Effective packet number for early stage internet traffic identification, Neurocomputing 156 (2015) 252-267.

[13] L. Bernaille, R. Teixeira, I. Akodkenou, A. Soule, K. Salamatian, Traffic classification on the fly, Coord. Chem. Rev. 36 (2) (2006) 23-26.

[14] N.-F. Huang, G.-Y. Jai, H.-C. Chao, Early identifying application traffic with appli-cation characteristics, in: Proceeding of the 2008 IEEE International Conference on Communications, IEEE, 2008, pp. 5788-5792.

[15] B. Hullár, S. Laki, A. Gyorgy, Early identification of peer-to-peer traffic, in: Pro-ceeding of the 2011 IEEE International Conference on Communications, IEEE, 2011, pp. 1-6.

[16] T.T. Nguyen, G. Armitage, P. Branch, S. Zander, Timely and continuous machine-learning-based classification for interactive ip traffic, IEEE/ACM Trans. Netw. 20 (6) (2012) 1880-1894.

[17] A. Rizzi, S. Colabrese, A. Baiocchi, Low complexity, high performance neuro-fuzzy system for internet traffic flows early classification, in: Proceeding of the 2013 9th International Wireless Communications and Mobile Computing Conference, IEEE, 2013, pp. 77-82.

[18] B. Pfülb, C. Hardegen, A. Gepperth, S. Rieger, A study of deep learning for network traffic data forecasting, in: Proceeding of the International Conference on Artificial Neural Networks, Springer, 2019, pp. 497-512.

[19] P. Poupart, Z. Chen, P. Jaini, F. Fung, H. Susanto, Y. Geng, L. Chen, K. Chen, H. Jin, Online flow size prediction for improved network routing, in: Proceeding of the 2016 IEEE 24th International Conference on Network Protocols, IEEE, 2016, pp. 1-6.

[20] X.J. Zhu, Semi-supervised learning literature survey, http://digital.library.wisc.edu/1793/60444, 2005.

[21] A. Blum, T. Mitchell,Combining labeled and unlabeled data with co-training, in:Proceedings of the Eleventh Annual Conference on Computational Learning Theory, 1998, pp. 92-100.

[22] Z.-H. Zhou, M. Li, Tri-training: exploiting unlabeled data using three classifiers, IEEE Trans. Knowl. Data Eng. 17 (11) (2005) 1529-1541.

[23] X. Ning, X. Wang, S. Xu, W. Cai, L. Zhang, L. Yu, W. Li, A review of research on co-training, Concurr.comput. (2021) e6276.

[24] H. He, X. Luo, F. Ma, C. Che, J. Wang, Network traffic classification based on en-semble learning and co-training, Sci. China Inf. Sci. 52 (2) (2009) 338-346.

[25] J. Yan, X. Yun, Z. Wu, H. Luo, S. Zhang, S. Jin, Z. Zhang, Online traffic classification based on co-training method, in: Proceeding of the 2012 13th International Confer-ence on Parallel and Distributed Computing, Applications and Technologies, IEEE, 2012, pp. 391-397.

[26] N. Seddigh, B. Nandy, D. Bennett, Y. Ren, S. Dolgikh, C. Zeidler, J. Knoetze, N.S. Muthyala, A framework & system for classification of encrypted network traffic us-ing machine learning, in: Proceeding of the 2019 15th International Conference on Network and Service Management, IEEE, 2019, pp. 1-5.

[27] C. Yu, J. Lan, J. Xie, Y. Hu,Qos-aware traffic classification architecture using ma-chine learning and deep packet inspection in sdns, Proc.comput. Sci. 131 (2018) 1209-1216.

[28] S. Zhao, Y. Zhang, P. Chang, Network traffic classification using tri-training based on statistical flow characteristics, in: Proceeding of the 2017 IEEE Trustcom/Big-DataSE/ICESS, IEEE, 2017, pp. 323-330.

[29] M. Alauthman, N. Aslam, M. Al-Kasassbeh, S. Khan, A. Al-Qerem, K.-K.R. Choo, An efficient reinforcement learning-based botnet detection approach, J. Netw. Comput. Appl. 150 (2020) 102479.

[30] D. Wu, B. Fang, J. Wang, Q. Liu, X. Cui,Evading machine learning botnet detection models via deep reinforcement learning, in: Proceeding of the ICC 2019-2019 IEEE International Conference on Communications, IEEE, 2019, pp. 1-6.

[31] M. Woodward,C. Finn, Active one-shot learning, arXiv preprint, arXiv:1702.06559.

[32] J. Janisch, T. Pevny` V. Lisy`,Classification with costly features using deep reinforce-ment learning, in:Proceedings of the AAAI Conference on Artificial Intelligence, vol. 33, 2019, pp. 3959-3966.

[33] T.-Y. Mu, A. Al-Fuqaha, K. Shuaib, F.M. Sallabi, J. Qadir, Sdn flow entry management using reinforcement learning, ACM Trans. Auton. Adapt. Syst. 13 (2) (2018) 1-23.

[34] W.-X. Liu, J. Cai, Y. Wang, Q.C. Chen, D. Tang, Mix-flow scheduling using deep reinforcement learning for software-defined data-center networks, Internet Technol. Lett. 2 (3) (2019) e99.

[35] J. Zhang, M. Ye, Z. Guo, C.-Y. Yen, H.J. Chao, Cfr-rl: traffic engineering with rein-forcement learning in sdn, IEEE J. Sel. Areas Commun. 38 (10) (2020) 2249-2259.

[36] S. Brooks, A. Gelman, G. Jones, X.-L. Meng, Handbook of Markov Chain Monte Carlo, first ed., Chapman & Hall CRC, Beijing, 2011.

[37] S.V. Kumar, L. Vanajakshi, Short-term traffic flow prediction using seasonal arima model with limited input data, Eur. Transp. Res. Rev. 7 (3) (2015) 1-9.

[38] A.R. Abdellah, O.A.K. Mahmood, A. Paramonov, A. Koucheryavy, Iot traffic predic-tion using multi-step ahead prediction with neural network, in: Proceeding of the 2019 11th International Congress on Ultra Modern Telecommunications and Con-trol Systems and Workshops, IEEE, 2019, pp. 1-4.

[39] H. Cheng, Z. Xie, Y. Shi, N. Xiong, Multi-step data prediction in wireless sensor networks based on one-dimensional cnn and bidirectional lstm, IEEE Access 7 (2019) 117883-117896.

[40] M. Lopez-Martin, B. Carro, A. Sanchez-Esguevillas, Iot type-of-traffic forecasting method based on gradient boosting neural networks, Future Gener. Comput. Syst. 105 (2020) 331-345.

[41] M. Balanici, S. Pachnicke, Classification and forecasting of real-time server traffic flows employing long short-term memory for hybrid e/o data center networks, J. Opt.commun. 13 (5) (2021) 85-93.