The road to autonomous broadband networks
If we were to build a broadband network today, with all the current technology at our disposal, what would it look like?
Top of mind for many would be an autonomous network that mostly looks after itself. That’s because operators currently face an overwhelming number of tasks to deliver the best possible broadband experience, while much of the service level expectations and network complexity could be managed or eliminated through automation and machine learning.
But we’re not starting from scratch. The vast investments made in current broadband infrastructure and the imperative of maintaining existing services mean that we can’t start over, nor make a sudden “big bang” leap into a fully automated future. But we do have the opportunity to start the transition.
For many years, the industry has used scripted, proprietary, network-centric systems with inflexible point-and-click management systems and FCAPS actions needing endless human intervention. But now, a software-defined access network (SDAN) enables us to apply more automated and intelligent control over the network. We’ve coalesced around the industry-wide adoption and standardization of open APIs for network management, such as NETCONF/YANG. Standardization and interoperability are fundamental to reaping the benefits of truly open and programmable networks. They save resources for both operators and vendors, as they are freed from endless integration and testing cycles and can focus instead on innovation and customer value. We’re just beginning to see these evolutions in our networks, but the automation is still very inflexible, along the lines of “if x happens, do y”.
In pushing the boundaries of what can be achieved through SDAN, Nokia is pioneering a new wave of intelligence in the fixed access network. Intent-based automation allows operators to define policies and let the network self-adjust, automatically finding misconfigurations and fixing them via audit facilities, synchronization processes and closed-loop automation. Intent-based networking is ideal for automating repetitive and dynamic processes that are a drain on a service provider’s resources.
Of course, like many areas of technology, the answer to full autonomics lies in artificial intelligence and machine learning (AI/ML). With AI/ML finding insight from the data lake and sending it to the network’s intent engine, Nokia brings operators one step closer to self-aware, self-governing networks that require little manual intervention for routine tasks. Intelligence + intent allows us to automate operations, predict and prevent issues, and provide detailed analysis for anomaly detection, action recommendation and network capacity planning. With mundane, day-to-day activities taken over by machines, operators’ human capital can focus on what it does best: serving human customers.
The ability of SDAN to create this autonomous future is very close. But let’s remember today’s reality: there is no intention of risking current services and every desire to protect the investment operators have made in their networks. Operators want to deploy SDAN in a pragmatic way that jeopardizes neither. That’s why Nokia has invested in a solution to help operators adopt SDAN at their own pace, always with an eye on creating immediate incremental revenue as well as operational improvements.
The road to autonomics in broadband networks may look long, but Nokia will get you there faster than you think.