스마트미디어저널 (Smart Media Journal)

  • 제13권6호
  • /
  • Pages.24-34
  • /
  • 2024
  • /
  • 2287-1322(pISSN)
  • /
  • 2288-9671(eISSN)
한국스마트미디어학회 (THE KOREAN INSTITUTE OF SMART MEDIA)
권호 표지
DOI QR코드

DOI QR Code

밀집 샘플링 기법을 이용한 네트워크 트래픽 예측 성능 향상

Improving prediction performance of network traffic using dense sampling technique

  • 이진선 (우석대학교 정보보안학과) ;
  • 오일석 (전북대학교 컴퓨터인공지능학부)
  • 투고 : 2024.02.28
  • 심사 : 2024.05.14
  • 발행 : 2024.06.28
PDF 다운로드
⟨ 이전 논문 다음 논문 ⟩

초록

시계열인 네트워크 트래픽 데이터로부터 미래를 예측할 수 있다면 효율적인 자원 배분, 악성 공격에 대한 예방, 에너지 절감 등의 효과를 거둘 수 있다. 통계 기법과 딥러닝 기법에 기반한 많은 모델이 제안되었는데, 이들 연구 대부분은 모델 구조와 학습 알고리즘을 개선하는 일에 치중하였다. 모델의 예측 성능을 높이는 또 다른 접근방법은 우수한 데이터를 확보하는 것이다. 이 논문은 우수한 데이터를 확보할 목적으로, 시계열 데이터를 증강하는 밀집 샘플링 기법을 네트워크 트래픽 예측 응용에 적용하고 성능 향상을 분석한다. 데이터셋으로는 네트워크 트래픽 분석에 널리 사용되는 UNSW-NB15를 사용한다. RMSE와 MAE, MAPE를 사용하여 성능을 분석한다. 성능 측정의 객관성을 높이기 위해 10번 실험을 수행하고 기존 희소 샘플링과 밀집 샘플링의 성능을 박스플롯으로 비교한다. 윈도우 크기와 수평선 계수를 변화시키며 성능을 비교한 결과 밀집 샘플링이 일관적으로 우수한 성능을 보였다.

If the future can be predicted from network traffic data, which is a time series, it can achieve effects such as efficient resource allocation, prevention of malicious attacks, and energy saving. Many models based on statistical and deep learning techniques have been proposed, and most of these studies have focused on improving model structures and learning algorithms. Another approach to improving the prediction performance of the model is to obtain a good-quality data. With the aim of obtaining a good-quality data, this paper applies a dense sampling technique that augments time series data to the application of network traffic prediction and analyzes the performance improvement. As a dataset, UNSW-NB15, which is widely used for network traffic analysis, is used. Performance is analyzed using RMSE, MAE, and MAPE. To increase the objectivity of performance measurement, experiment is performed independently 10 times and the performance of existing sparse sampling and dense sampling is compared as a box plot. As a result of comparing the performance by changing the window size and the horizon factor, dense sampling consistently showed a better performance.

키워드

참고문헌

  1. .F. Torres et al" "Deep learning for time-series forecasting: a survey," Big Data, Vol. 9, No. 1, pp. 3-21, 2021. https://doi.org/10.1089/big.2020.0159
  2. S. Meisenbacher et al., "Review of automated time series forecasting pipelines," WIREs Data Mining and Knowledge Discovery, 2022.
  3. Mazaher Kanpour and Shahid Raza, "More than malware: unmasking the hidden risk of cybersecurity regulations," International Cybersecurity Law Review, Vol. 5, pp. 169-212, 2024. https://doi.org/10.1365/s43439-024-00111-7
  4. Dexian Yang, Jiong Yu, Xusheng Du, Zhenzhen He, and Ping Li, "Energy saving strategy of cloud data compiling based on convolutional neural network and policy gradient algorithm, PLoS ONE, Vol. 17, No. 12, pp. 1-18, 2022. https://doi.org/10.1371/journal.pone.0279649
  5. Aaron Chen, Jeffrey Law, and Michal Aibin, "A survey on traffic prediction techniques using artificial intelligence for communication networks," Telecom, Vol. 2, pp. 518-535, 2021. https://doi.org/10.3390/telecom2040029
  6. Gabriel O. Ferreira et al., "Forecasting network traffic: a survey and tutorial with open-source comparative evaluation," IEEE Access, Vol. 11, pp. 6018-6044, 2023. https://doi.org/10.1109/ACCESS.2023.3236261
  7. Mingle Xu, Sook Yoon, Alvaro Fuentes, and Dong Sun Park, "A comprehensive survey of image augmentation techniques for deep learning," Pattern Recognition, Vol. 137, 2023.
  8. Qingsong Wen et al., "Time series data augmentation for deep learning: a survey," Proceedings of International Joint Conference on Artificial Intelligence, pp. 4653-4660, Montreal, Canada, 2021.
  9. Zijun Gao, Lingbo Li, and Tianhua Xu, "Data augmentation for time-series classfication: an extensive empirical study and comprehensive survey," arXiv:2310.10060v2, 2023.
  10. Z. Cui, Wenlin Chen, and Yixin Chen, "Muiti-scale neural networks for time series classification," arXiv:1603.06995v4, 2016.
  11. Bin Qian et al., "Dynamic multi-scale convolutional neural netwotkr for time series classification," IEEE Access, Vol. 8, pp. 109732-109746, 2020. https://doi.org/10.1109/ACCESS.2020.3002095
  12. Arthur Le Guennec, Simon Malinowski, and Romain Tavenard, "Data augmentation for time series classification using convolutional neural networks," ECML/PKDD Workshop on Advanced Analytics and Learning on Temporal Data, Riva Del Garda, Italy, 2016.
  13. Wenbo Yang, Jidong Yuan, and Xiaokang Wang, "SFCC: data augmentation with stratified Fourier coeeficients combination for time series classification," Neural Processing Letters, Vol. 55, pp. 1833-1846, 2023. https://doi.org/10.1007/s11063-022-10965-9
  14. I.-S. Oh and J. S. Lee, "Dense sampling of time series for forecasting," IEEE Access, Vol. 10, pp. 75571-75580, 2022. https://doi.org/10.1109/ACCESS.2022.3191668
  15. Nour Moustafa and Jill Slay, "UNSW-NB15: a comprehensive data set for network intrusion detection systems," Military Communications and Information Systems Conference, pp. 1-6, Canberra, ACT, Australia, Nov. 2015.
  16. Sangjoon Jung, Chonggun Km, and Younky Chung, "A prediction method of network traffic using time series models," International Conference on Computational Science and Its Applications, Vol. 3, pp. 234-243, 2006.
  17. B. Vujicic, Hao Chen, and L. Trajkovic, "Prediction of traffic in a public safety network," IEEE International Symposium on Circuits and Systems, Island of Kos, 2006.
  18. Mingyang Zhang, Haohao Fu, Yong Li, Sheng Chen, "'Understanding urban dynamics from massive mobile traffic data," IEEE Transactions on Big Data, Vol. 5, No. 2, pp. 266-278, 2019. https://doi.org/10.1109/TBDATA.2017.2778721
  19. Nicolai Stepanov, Daria Alekseeva, Aleksandr Ometov, and Elena Simona Lohan, "Applying machine learning to LTE traffic prediction: comparison of bagging, random forest, and SVM," International Congress on Ultra Modem Telecommunications and Control Systems and Workshops, pp. 119-123, Brno, Czech Republic, 2020.
  20. Zhongda Tian, "Network traffic prediction method based on wavelet transform and multiple models fusion," International Journal of Communication Systems, Vol. 33, no. 11, 2020.
  21. Timothy Chadza, Konstantinos G. Kyriakopoulos, and Sangarapillai Lambotharan, "Contemporary sequential network attacks prediction using hidden Markov model," Internationa Conference on Privacy, Security and Trust, pp. 1-3, Fredericton, NB, Canada, 2019.
  22. Melinda Barabas, Georgeta Boanea, Andrei B. Rus, Virgil Dobrota, and Jordi Domingo-Pascuak, "Evaluation of network traffic prediction based on neural networks with multi-task learning and multiresolution decomposition," International Conference on Intelligent Computer Communication and Processing, pp. 95-102, Cluj-Napoca, Romania, 2011.
  23. Abdelhadi Azzouni and Guy Pujolle, "A long short-term memory recurrent neural network framework for network traffic matrix prediction," arXiv:1705.05690v3, 2017.
  24. Haipeng Lu and Fan Yang, "A network traffic prediction model based on wavelet transformation and LSTM network," IEEE International Conference on Software Engineering and Service Science, pp. 1-4, Beijing, China, 2018.
  25. Nipun Ramakrishnan and Tarun Soni, "Network traffic prediction using recurrent neural network," IEEE International Conference on Machine Learning and Applications, pp. 187-193, Orlando, FL, USA, 2018.
  26. Jing Wang et al., "Spatiotemporal modeling and prediction in cellular networks: a big data enabled learning approach," IEEE Conference on Computer Communications, pp. 1-9, Atlanta, GA, USA, 2017.
  27. Xianglong Luo, Danyang Li, Yu Yang, and Shengrui Zhang, "Spatiotemporal traffic flow prediction with kNN and LSTM," Journal of Advanced Transportation, pp. 1-10, 2019.
  28. Dario Bega, Marco Gramaglia, Marco Fiore, Albert Banchs, and Xavier Costa-Perez, "DeepCog: cognitive network management in sliced 5G networks with deep learning," IEEE Conference on Computer Communications, pp. 280-288, Paris, France, 2019.
  29. Hanyu Yang et al., "A network traffic forecasting method based on SA optimized ARIMA-BP neural network," Computer Networks, Vol. 193, 2021.
  30. Zi Wang et al., "Spatio-temporal cellular traffic prediction for 5G and beyond: a graph neural networks-based approach," IEEE Transactions on Industrial Informatics, Vol. 19, No. 4, pp.5722-5731, 2023. https://doi.org/10.1109/TII.2022.3182768
  31. Xu Zhou et al., "Large-scale cellular traffic predication based on graph convolutional networks with transfer learning," Neural Computing and Applications, Vol. 34, pp. 5549-5559, 2022. https://doi.org/10.1007/s00521-021-06708-x
  32. Leonardo Lo Shiavo, Marco Fiore, Marco Gramaglia, Albert Banchs, and Xavier Costa-Perez, "Forecasting for network management with joint statistical modeling and machine learning," IEEE International Symposium on a World of Wireless, Mobile, and Multimedia Networks, pp. 60-69, Belfast, United Kingdom, 2022.
  33. Jian Su, Huimin Cai, Zhengguo Sheng, A.X. Liu, and Abdullah Baz, "Traffic prediction for 5G: a deep learning approach based on lightweight hybrid attention networks," Digital Signal Processing, Vol. 146, 2024.
  34. Aston Zhang, Zachary C. Lipton, Mu Li, and Alexander J. Smola, Dive into Deep Learning, Cambridge University Press, 2023.