How digital twins improve internet speeds
In this blog series, we’ve been looking at how Artificial Intelligence (AI) and Machine Learning (ML) are transforming operations in fiber broadband networks that can sense, think and act. In this blog, we’re taking a look at the application of digital twins.
When internet speeds aren’t as advertised
What you see isn’t always what you get when it comes to internet speeds. While a user may subscribe to a 200 Mb/s service, in reality the speed will often fluctuate during the day, depending on what’s going on in the network. On 8th November 2023, for example, the highly anticipated release of Call of Duty: Modern Warfare III created a massive 25.1 Tbps spike in Virgin Media’s UK network. That kind of activity can cause speeds to drop dramatically during peak hours. And with roughly 1.5 billion PC and console gamers in the world1, and around 5,000 new games released on the major platforms each year2, congestion is a daily risk.
This can be frustrating for subscribers and a concern for operators, as regulators are developing monitoring schemes to measure broadband performance and better protect consumers. The Federal Communications Commission in the U.S. has defined a testing framework to confirm that services subsidized by the Connect America Fund (CAF) meet basic performance standards. The FCC mandates that at least 80 percent of network speed measurements must be at 80 percent of required speeds.
Scoring well on these metrics puts operators in a strong position in the market, as consumers like the assurance that their internet connection meets advertised performance levels.
So, it’s not something you want to leave to chance.
Optimizing performance with digital twins
To take luck out of the equation, operators leave a lot of headroom in the fiber access network, which ensures that, during peak hours, at least one PON user can download at full speed. This approach is simple yet limited and overly conservative. It means PON networks are largely underutilized, tying up capital and resources without delivering full value.
Enter the digital twin and a far superior way of optimizing network capacity.
A network digital twin is broadly understood as a digital representation of a real-world network, with which it is synchronized. Digital twins represent the next step in network simulation software: using real-time linkages to physical network entities for monitoring and control, integrating big data analytics and AI, and offering flexible user interactions and what-if predictions.
Digital twins have the potential to revolutionize network capabilities and operations. This can improve planning decisions, validate configurations, speed up troubleshooting, and help decide on corrective actions. This paper by Nokia Bell Labs explains all the benefits of digital twins and here we’ll look at how a digital twin can improve both the strategic and operational aspects of PON capacity management.
Enhancing PON capacity planning
Digital twin-based capacity planning predicts congestion risks much more accurately and is more focused on Quality of Experience (QoE) than traditional capacity planning. It avoids overly conservative service definitions and PON configurations, allowing network operators to increase PON utilization - with up to 33 percent higher peak rates or doubling the number of users - without increasing the risk of failed speed tests.
For strategic capacity planning, the digital twin enables what-if analysis. What is the impact if we increase the split ratio? What if a (multi-)gigabit service is introduced? The digital twin can also assess the impact of future traffic growth and determine when to upgrade to a new PON technology.
For operational capacity management, the digital twin monitors congestion levels and calculates the speed test success probability for all PONs in the network, eliminating the need for real speed tests that are disruptive. This enables operators to verify whether adding a user to the PON or upgrading a user to a higher service tier is feasible without jeopardizing speed-test success (and fulfilling the 80 percent-80 percent requirement). The digital twin can do more than pre-validate a service request; it can even suggest on which PON to add a user or predict the highest speed that can be offered without risking PON congestion.
In summary, the digital twin provides operators with valuable insights into their PON networks, enabling them to make informed decisions about capacity planning and management. This results in increased PON utilization, improved network performance, and a better user experience.
Be sure to download the white paper to understand all the ways that AI/ML are transforming fiber network operations. Also, why not check out the Nokia Altiplano Access Controller, the domain controller that brings sophisticated intelligence into the fixed access domain, which was awarded the #1 network automation platform for broadband access networks worldwide in 2024.
Sources
