- Malaysia’s Digital Nasional Berhad's 5G-Advanced network received TM Forum’s Level 4 autonomous network validation
- At Level 4, the network uses AI-based predictive analytics to anticipate issues and self-adjust for optimal performance
- The validation sets a precedent for other operators
Telcos across all geographies are under pressure to improve customer experience while bringing down their operational expenses — and it's one of the key reasons why they are keen to transition to autonomous networks, since that will help in reducing expenditure while ensuring optimal network performance. However, they face several challenges in this journey.
As per the recent STL Partners report, supported by RedHat and Capgemini, only 28% of the telcos expect to achieve high autonomous network maturity over the next three years.
One of the biggest challenges is legacy systems, which makes the adoption of AI tools and systems complex. In addition, telcos still work in silos, which means that data, both structured and unstructured, is spread across several departments, including networks, IT and customer service, making it extremely difficult to use it to train AI systems.
Another crucial problem is the budgetary constraints since the return on investment for such projects is not immediate and is spread over a long period. The telcos may also face people issues, including finding the right talent as well as resistance from the workforce in adopting AI technology.
Malaysia's DNB goes for Level 4 automation
Those challenges didn't stop Malaysia's state-owned wholesale 5G network operator Digital Nasional Berhad (DNB). Together with Ericsson, DNB achieved validation of TM Forum’s Level 4 autonomy in Service Assurance.
Malaysia established DNB to build and manage the country’s 5G infrastructure. It's network is used by several service providers, such as CelcomDigi, Maxis, YTL and U Mobile, to provide high-speed 5G experience to their subscribers.
Earlier this year, DNB partnered with Ericsson to build its 5G-Advanced (5G-A) network across the country. The wholesale operator's Throughput Management Capability system, which was validated under TM Forum’s Autonomous Network Level Assessment Validation (ANLAV) program, has been active for two years. It uses GenAI to discover as well as resolve issues automatically, to ensure a reliable and fast 5G network. It also leverages digital twins to test resolutions before applying them to the network.
Level 4 key factors
At Level 4, the network leverages AI-based predictive analytics to forecast issues as well as to adjust to maintain optimal performance. The network uses Ericsson’s AI-enabled Intent-based Operations (IBO) and can support industrial and enterprise-grade services, according to a press release from Ericsson.
"DNB’s Level 4 validation is significant for two reasons. First, it proves that AI-driven, intent-based automation can work at commercial scale in a live 5G network - not just in lab conditions. Second, this sets a benchmark for wholesale network models globally. DNB leapfrogged legacy operators by building autonomy into its 5G infrastructure from the outset, rather than retrofitting it – a strategic advantage that other state-backed or greenfield operators should study closely,” said Arun Menon, Principal Analyst at MTN Consulting.
Fierce Network reached out to DNB for comments for this article but did not receive a reply.
“Autonomous Network has been one of the core design principles of DNB’s 5G rollout from the very beginning. By integrating automation and AI capabilities as the technology matured, we’ve built a highly efficient network that delivers world-class 5G connectivity at one of the lowest costs globally,” said Ken Tan, Chief Technology Officer (CTO) at DNB in the press release about the news.
“Achieving Level 4 autonomy in such an important area boosts the ability of Malaysia’s Network Operators to offer its customers new and valuable differentiated connectivity services, based on the very latest advances in autonomous networks,” said Bradley Mead, Head of Managed Network Services at Ericsson.
Commenting on the key factors required to reach Level 4 or Level 5 of network autonomy, Menon of MTN Consulting said, “For reaching Level 4 or 5, three factors are non-negotiable: deep AI/ML integration for closed-loop, predictive operations; adoption of TM Forum's Open APIs and modular architectures to enable cross-vendor automation without vendor lock-in; and clean, high-quality data pipelines. The operators advancing fastest are those treating autonomy as a business transformation, not just a technology upgrade.”