Page 127 - ITU Journal Future and evolving technologies Volume 2 (2021), Issue 4 – AI and machine learning solutions in 5G and future networks

ITU Journal on Future and Evolving Technologies, Volume 2 (2021), Issue 4

one prediction for each learned model. Then we add a [9] Irene Katzela and Mischa Schwartz, “Schemes
new model which learns from the intermediate for fault identi ication in communication networks,”
predictions of the same target. IEEE/ACM Transactions on networking, vol. 3, no. 6,
pp. 753–764, 1995.
6. FUTURE WORK
[10] Anastasios T Bouloutas, George W Hart, and Mischa
As for future work, we are aiming to compare different Schwartz, “Fault ication using a inite state
machine learning methods, for example, DT, MLP, RF, XG‑ machine model with unreliable partially observed
Boost, etc. with different data sets, not only KDDI’s, to ex‑ data sequences,” IEEE Transactions on Communica‑
plore clearer boundaries of the most suitable range be‑ tions, vol. 41, no. 7, pp. 1074–1083, 1993.
tween each method so that we could take the optimal
[11] Cynthia S Hood and Chuanyi Ji, “Proactive network‑
solution for diverse failures. In addition, analyzing the
fault detection [telecommunications],” IEEE Trans‑
topology and applying graphical convolution networks at
actions on reliability, vol. 46, no. 3, pp. 333–341,
the stage of determining the importance of features
might prove a promising result. 1997.
[12] Eleftheria Athanasopoulou and Christoforos N Had‑
REFERENCES jicostis, “Probabilistic approaches to fault detection
in networked discrete event systems,” IEEE trans‑
[1] “Coronavirus: Five irms booming despite the actions on neural networks, vol. 16, no. 5, pp. 1042–
lockdown,” IV , {https://www.bbc.com/news/b 1052, 2005.
usiness-52383193}.
[13] I lhan Firat Kilinçer, Fatih Ertam, Orhan Yaman, and
[2] “How zoom became so popular during social dis‑ Ayhan Akbal, “Automatic fault detection with bayes
tancing,” IV , {https://www.cnbc.com/2020/ method in university campus network,” in 2017 In‑
04/03/how-zoom-rose-to-the-top-during-the ternational icial Intelligence and Data Process‑
-coronavirus-pandemic.html}. ing Symposium (IDAP). IEEE, 2017, pp. 1–4.

[3] “Google lost $1.7m in ad revenue during [14] Marc Ruiz, Francesco Fresi, Alba P Vela, Gianluca
youtube outage, expert says,” IV , {https: Meloni, Nicola Sambo, Filippo Cugini, Luca Poti, Luis
//finance.yahoo.com/news/google-lost-1-7m
Velasco, and Piero Castoldi, “Service‑triggered fail‑
-ad-144303640.html}.
ure ication/localization through monitoring
of multiple parameters,” in ECOC 2016; 42nd Eu‑
[4] Junichi Kawasaki, Genichi Mouri, and YusukeSuzuki, ropean Conference on Optical Communication. VDE,
“Comparative analysis of network fault classi ica‑ 2016, pp. 1–3.
tion using machine learning,” in NOMS 2020‑
2020 IEEE/IFIP Network Operations and Manage‑
[15] Carla Sauvanaud, Kahina Lazri, Mohamed Kaâniche,
ment Symposium. IEEE, 2020, pp. 1–6.
and Karama Kanoun, “Anomaly detection and root
cause localization in virtual network functions,” in
[5] Tianqi Chen and Carlos Guestrin, “Xgboost: A scal‑ 2016 IEEE 27th International Symposium on Soft‑
able tree boosting system,” in Proceedings of the
22nd acm sigkdd international conference on knowl‑ ware Reliability Engineering (ISSRE). IEEE, 2016,
pp. 196–206.
edge discovery and data mining, 2016, pp. 785–794.
[16] Srinikethan Madapuzi Srinivasan, Tram Truong‑
[6] Guolin Ke, Qi Meng, Thomas Finley, Taifeng Wang,
Huu, and Mohan Gurusamy, “Machine learning‑
Wei Chen, Weidong Ma, Qiwei Ye, and Tie‑Yan Liu,
“Lightgbm: A highly ef icient gradient boosting deci‑ based link fault ication and localization in
sion tree,” Advances in neural information processing complex networks,” IEEE Internet of Things Journal,
systems, vol. 30, pp. 3146–3154, 2017. vol. 6, no. 4, pp. 6556–6566, 2019.
[17] Rolf Stadler, Rafael Pasquini, and Viktoria Fodor,
[7] Mitchell Vásquez‑Bermúdez, Jorge Hidalgo, Marı́a “Learning from network device statistics,” Journal
del Pilar Avilés‑Vera, José Sánchez‑Cercado, and
Christian Roberto Antón‑Cedeño, “Analysis of a of Network and Systems Management, vol. 25, no. 4,
network fault detection system to support decision pp. 672–698, 2017.
making,” in International Conference on Technolo‑ [18] Karwan Qader, Mo Adda, and Mouhammd Al‑
gies and Innovation. Springer, 2017, pp. 72–83.
Kasassbeh, “Comparative analysis of clustering
techniques in network ic faults classi ication,”
[8] Todd E Marques, “A symptom‑driven expert system International Journal of Innovative Research in Com‑
for isolating and correcting network faults,” IEEE puter and Communication Engineering, vol. 5, no. 4,
Communications Magazine, vol. 26, no. 3, pp. 6–13, pp. 6551–6563, 2017.
1988.

© International Telecommunication Union, 2021 111