To keep up with growing network complexity, telcos are applying automation techniques enabled by AI and ML in as many network operations as possible
As 5G continues to evolve, it also continues to become more complex. Virtualization and disaggregation are happening in tandem with deployment of network workloads in hybrid cloud environments, including distributed clouds deep in the network at radio sites or even at customer premises. As a result, configuring, provisioning and assuring networks through manual — or even the standard automation strategies that telcos have been relying on for years — is no longer possible, from an economic or practical perspective. Therefore, applying automation techniques enabled by artificial intelligence (AI) and machine learning (ML) in as many network operations as possible — referred to as AIOps, or Artificial Intelligence for IT Operations — is emerging as the only way for telcos to keep up.
Against this dynamic, complex backdrop, lies the criticality of meaningful data. “When we think about 5G, edge computing and IoT, the common pattern in all of these cases is that the volume and velocity of the data grows exponentially,” explained Javier Antich Romaguera, senior director of product management at Selector AI, which provides operational intelligence of multi-cloud infrastructure and performance-sensitive managed services. “With these technology evolutions, network infrastructure is larger and more complex than ever. Further, cloud-native means that infrastructure is composed of thousands or hundreds of thousands of tiny components… [that] are working together and are more likely to fail or have deviations.”
This creates a particularly challenging environment for the network operator in charge of managing all of these moving pieces, as well as ensuring they extract the right intelligence or information from the massive amount of data being generated. A workforce shortage resulting from economic pressures and an expanding skills gap only adds another roadblock to keeping up with the complexity of networks and the data they produce.
And then, even as these hurdles grow, the infrastructure itself is becoming more critical, meaning the cost of any network failure is exponentially higher than before because more and more businesses, people and operations depend on this connected infrastructure.
“When you combine these three aspects — networks are more complex than ever, there are fewer resources than ever, and infrastructure is more critical than ever — this is a perfect storm and the only way out is with data and algorithms,” Romaguera claimed.
Another factor at play here, according to Per Kangru, technologist in the CTO Office at Viavi Solutions, is the move towards Open RAN, which seeks to disaggregate RAN hardware and software on vendor-neutral platforms is provided the necessary organization motivation for telcos to commit to AIOps.
“Open RAN moves telcos into an area where everything they historically knew about network management is now gone; they now have new offerings that are completely different architecturally and in how they are managed,” said Kangru, arguing that such a massive transformation begets true organizational innovation, opening the door for things like AIOps. “When you go to Open RAN, you’re forcing people to think differently,” he claimed. That open-mindedness, in and of itself, makes introducing AI and automation deeper into an organization’s operations more feasible. Beyond that, however, Kangru acknowledged that O-RAN doesn’t really offer anything else that makes it more or less conducive to automation than previous architectures.
“AI drives a new culture based upon building an environment with much more agility and a higher ability to use resources in order to deliver the right capability at the right time,” he said, adding that being successful in this space comes down to having “an extremely strong idea about what your business is and what you want to do and then driving that through the organization.” Without these elements, he cautioned, AIOps initiatives will “very, very difficult” to achieve.
Keep an eye out for the complete report about this topic coming soon called Bringing AIOps to telecom: When will operators take their hands off the wheel?
