A new ITU Focus Group will establish a basis for ITU standardization to assist machine learning in bringing more automation and intelligence to ICT network design and management, particularly for 5G networks.
The creation of the Focus Group was agreed by ITU’s standardization expert group for future networks, ITU-T Study Group 13, at its meeting at ITU headquarters in Geneva last week.
Machine-learning algorithms are helping operators to make smarter use of network-generated data. These algorithms enable ICT networks and their components to adapt their behavior autonomously in the interests of efficiency, security and optimal user experience.
The ITU Focus Group on Machine Learning for Future Networks including 5G will lead an intensive one-year investigation into where technical standardization could support emerging applications of machine learning in fields such as big data analytics, network management and orchestration, and security and data protection.
“ITU Focus Groups define new directions in ITU standardization,” says Chaesub Lee, director of the ITU Telecommunication Standardization Bureau. “These groups deliver base documents to stimulate international standardization work. They are effective in accelerating ITU studies in fields of growing strategic relevance to the ITU membership.”
“Machine learning and artificial intelligence are finding promising applications in communications networking,” says the Focus Group’s Chairman, Slawomir Stanczak of Germany’s Fraunhofer Heinrich-Hertz-Institut. “This Focus Group will establish a basis for ITU standards experts to capitalize on machine learning in their preparations for the 5G era.”
The first meeting of the Focus Group is scheduled for January 29-February 2, 2018. Contributions are invited on state-of-the-art use cases of machine learning and underlying technical requirements.
An analysis of emerging use cases will inform the Focus Group’s development of technical specifications to meet the requirements of such use cases with respect to network architectures, interfaces, protocols, algorithms and data formats.
The Focus Group will consider machine-learning methods’ compatibility with a wide variety of fixed and mobile communication stacks, encouraging the development of methods attuned to the operational requirements of the networking industry.
Interoperability is high on the agenda. The Focus Group will propose means to train, adapt, compress and exchange machine-learning algorithms. This work will promote the emergence of an ecosystem able to support the interaction of multiple machine-learning algorithms. Participation is open to all interested parties.