From data-driven to intent-based network operations
This is the story of how Malaysia’s Digital Nasional Berhad used Ericsson Operations Engine to adopt automated and data-driven processes that helped them build the world’s first multi-operator core network, before successfully trialling intent-based operations (IBOs) that unlock greater profitability.
Highlights
Digital Nasional Berhad created the world’s first multi-operator core network, which met the needs of six different telecom service providers and ensured affordable 5G connectivity for the people of Malaysia, just six months after its founding, thanks to Ericsson Operations Engine.
Automation was used to manage the complexity of these networks so effectively, that significant increases to network uptime and reductions to operating costs and alarm rates were achieved. This success led to DNB being chosen to trial the use of IBOs on its network, which, in turn, led to the introduction of monetization, expanded business outcomes and an open route to differentiated services with distinct slices and guaranteed SLAs.
DNB’s challenges
- each service provider had their own complex networks, the complexity of which only increased over time
- DNB had to manage six separate networks, and six distinct sets of systems and processes
- DNB had to ensure that the SLA requirements of each network was met without any conflict
- software upgrades and new rollouts had to maintain total network availability and integrity
- a multi-operator core network had never been built before
- DNB were a startup who had to begin from the ground up
The solution
To manage the challenges it faced during the rollout of the network, DNB turned to Ericsson Operations Engine (EOE). This operating model is used to run multi-vendor, multi-technology network operations and facilitate the transformation to autonomy while ensuring high customer experience. EOE replicates and improves upon the reasoning of skilled engineers, by making use of closed-loop AI-based automation capabilities and data-driven processes that utilize machine learning to automate a network’s management, as well as cognitive core-based, intent-based automation and API-based connectivity solutions while remaining capable of resolving conflicting requirements without compromising on service quality or profitability.
Qualified personnel remain able to provide important oversight and actionable insights are obtained from advanced analytics, which help operations move from a reactive stance to a predictive one over time, thanks to the intelligence gleaned from machine learning. Additionally, standardized global processes reduce complexity and increase scalability, while the combination of network engineering and data science allows for the introduction of new processes that help networks run more efficiently.
Ericsson Operations Engine
There was an increase in efficiency with a reduction in downtime and alarm rate, while there was a significant network performance improvement. Customer experience improved overall, as did the service desk response time, as auto-actuation of corrective measures meant that customer complaint resolution time fell by 90 percent and the proportion of automatically created trouble tickets increased to 95 percent. Auto-analysis and correlation of alarms and actuation in the network led to a 500 percent reduction in alarm count 6 months after its introduction. Feedback loops from DNB’s network data combined with analytical models brought about a network uptime greater than 99.8 percent.
The management of DNB’s wholesale operation was set up with an inbuilt data-driven process that led to the achievement of measurable business outcomes, while real-time network visualization capabilities provided by autonomous analytical dashboards granted superior network visibility for customer experience management. The key to DNB’s success was the fact that the network was system driven with some assistance from humans.
The result: a step toward network autonomy
- The successful proof of concept for IBOs introduced a pathway to profitability, and to affordably introducing additional 5G capabilities in a way that scales with growth.
- Replicating and automating engineer decision-making will unlock service diversification and variability at scale.
- Simplifying processes and managing ever greater complexity through EOE automation brings in performance enhancement that cannot be matched by human involvement.
