Page 96 - Proceedings of the 2018 ITU Kaleidoscope

‎ 2018 ITU Kaleidoscope Academic Conference‎

developing standards for automation of network operation [2] Recommendation ITU-T Y.3102 (2018) “Framework of IMT-
control and management functions. 2020 network”.
[3] R. Li et al., “Intelligent 5G: When cellular networks meet
artificial intelligence,” IEEE Communications Magazine, Vol.
5.3 ISO/IEC JTC 1/SC42 24, No. 5, pp. 175-183, Mar. 2017.
ISO/IEC JTC 1/SC42 Artificial Intelligence [15] has been [4] M.G. Kibria et al., “Big data analytics, machine learning and
created in October 2017 to serve as the focus and artificial intelligence in next-generation wireless networks,”
proponent for ISO/IEC JTC 1’s standardization program on IEEE Access (Early Access), May 2018.
artificial intelligence, and guide JTC 1, IEC, and ISO [5] T. Miyazawa, M. Jibiki, V.P. Kafle, and H. Harai,
committees for the development of AI applications. It has “Autonomic resource arbitration and service-continuable
network function migration along service function chains,”
been developing several ISO standards on big data and AI, IEEE/IFIP Network Operations and Management Symposium
such as big data reference architecture, AI concepts and (NOMS2018), Apr. 2018.
terminology, and framework for AI systems using machine [6] K. Katsuura, M. Miyauchi, T. Numazaki, Y. Kurogouchi, Y.
learning. It has not yet started developing documents on the Satoh, and T. Koseki, “IaaS Automated operations
application of AI or machine learning techniques on management solutions that improve virtual environment
networks. efficiency,” NEC Technical Journal, vol.8, no.2, Apr. 2014,
pp. 29-32.
[7] J. Qadir et al., “Artificial intelligence enabled networking,”
5.4 TM Forum Smart BPM IEEE Access, vol. 3, 2015, pp. 3079-3082.
TM Forum’s Catalyst Project on Smart BPM (Business [8] X. Wang, X. Li, and V.C.M. Leung, “Artificial intelligence
Process Management) is investigating the applicability of based techniques for emerging heterogeneous network: State
AI-based decision modeling in telecom business processes of the arts, opportunities and challenges,” IEEE Access, vol.
such as resource provisioning, fault management, QoS 3, 2015, pp. 1379-91.
assurance, and customer management [16]. AI support [9] C. Jiang et al., “Machine learning paradigms for next-
generation wireless networks,” IEEE Wireless Commun., vol.
makes the workflow system capable of reacting to 24, no. 2, Apr. 2017, pp. 98-105.
exceptional conditions in the network service lifecycle [10] E. Alpaydm, Introduction to Machine Learning, Third
orchestration that includes planning, delivery, deployment, Edition, MIT Press, 2014.
and operation functionalities. In their approach, the AI [11] T.E. Bogale, X. Wang, and L. Le, “Machine intelligence
support system works as a secretary to recommend techniques for next-generation context-aware wireless
appropriate actions for handling exceptions. They mainly networks,” ITU Journal: ICT Discoveries, Special Issue No. 1,
consider using Multi-label Deep Neural Network in AI Feb. 2018.
support system, which takes input data from the [12] N. Kato et al., “The deep learning vision for heterogeneous
orchestrator, alarm management system, and other network network traffic control: proposal, challenges and future
functions. TM Forum has demonstrated the concept model perspective,” IEEE Wireless Communications, vol. 24, no. 3,
June 2017, pp. 146-153,
for the interaction between AI and human for network fault [13] Focus Group on Machine Learning for Future Networks
detection and recovery on the basis of training data including 5G, https://www.itu.int/en/ITU-
including operator’s feedback. T/focusgroups/ml5g/Pages/default.aspx, Visited 22 June 2018.
[14] ETSI ISG Experiential Network Intelligence (ENI),
6. CONCLUSION http://www.etsi.org/technologies-
clusters/technologies/experiential-networked-intelligence,
In this paper, we presented the 5G network slicing Visited 22 June 2018.
scenarios and list of tasks for the automation of slicing [15] ISO/IEC JTC 1/SC42 Artificial Intelligence,
functions. We then discussed some machine learning https://www.iso.org/committee/6794475.html, Visited 22
techniques that support for the automation of network June 2018.
functions. We also gave the overview of AI and machine [16] TM Forum Catalyst Project: Artificial Intelligence makes
learning related activities being carried out in various Smart BPM Smarter,
standards development organizations and industrial forums. https://www.tmforum.org/catalysts/smart-bpm, Visited 22
June 2018.
In future work, we extend this work by detail investigation
of a few representative machine learning techniques for
making autonomic decision to allocate and adjust the
computing and network resources of network slices on the
basis of service requirements and time-varying network
workloads. We will also bring this research outcome
gradually to ITU and other SDOs for standardization.

REFERENCES
[1] Recommendation ITU-R M.2083-0 (2015) “IMT Vision –
Framework and overall objectives of the future development
of IMT for 2020 and beyond”.

– 80 –