Page 33 - AI Ready – Analysis Towards a Standardized Readiness Framework

AI Ready – Analysis Towards a Standardized Readiness Framework

4.8 Future Networks and Telecommunications

As future network technologies such as IMT-2030 [97] evolve in parallel to evolutions in the AI
domain, integration of AI in future networks is shaping the nature of 6G and leading to new
applications and services. Topics of AI for Networks and Networks for AI have been discussed
which looks at the application of AI in networks as well as connectivity as an enabler for new
AI-based services. This section describes a typical use case where the application of the latest
advances in the field of Natural Language Processing (NLP), generative AI, and digital twins
could be used to enable self-optimization and integration of intelligent applications in networks.

4�8�1 Easing Operations Using AI/ML Solution for 6G and Beyond Network
Orchestration and Secure Network Operations

This use case [93] proposes to use Natural Language Processing (NLP), Digital Twins, and
Extended Reality (XR) to improve the network operations and orchestration in 5G/6G networks.
The solution simplifies the deployment and configuration of network functions (NFs). Operators
can input commands in natural language, transcending language barriers and reducing the
need for specialized technical knowledge. There are 3 aspects to the solution

A) A digital twin of the Kubernetes cluster simulates and predicts resource utilization, helping
operators manage infrastructure efficiently. This reduces the need for overprovisioning,
thereby saving costs and minimizing carbon emissions.
B) XR provides responsive and interactive observability, allowing operators to visualize and
interact with network configurations and projected traffic. This enhances decision-making
and troubleshooting capabilities. The solution continuously monitors network traffic for
anomalies using ML algorithms.
C) AI modules enforce appropriate security policies in real-time, and simulated attacks on
the digital twin help identify and fix vulnerabilities.

This use case addresses the problem of lack of technical expertise for deployment, operations,
and security. The AI Model trained in this solution generates Network intent with GenAI, ML/
AI, and deploys network slice. This is continuously monitored, and automatically scaled, and
the issues are self-healed. Traffic patterns are monitored, anomalies are detected, and security
policies are applied accordingly.

4.9 Accessibility

AI-based applications can significantly improve the lives of people with disabilities. The use cases
in this section use audio or visual processing designed to assist the communication process
for people with hearing or visual impairments. Generalization capabilities of local models and
solutions using standard approaches would scale the benefits to different parts of the world.

4�9�1 Machine Translations for Khmer Braille

This use case [43] aims to provide accessibility and facilitate communication for visually impaired
individuals in the Khmer Language. The machine translation system for Khmer Braille is publicly
available and free to access to the community. The use case collects text data in both Khmer
and Braille. The system is particularly targeting educational content so that equitable education
can be provided for individuals. To satisfy the needs, specific fine tuning for math and physics

26