Network automation is applicable to the optical layer and it is producing real benefits today. That was a key message that emerged from the Building Automated and Programmable Optical Networks session at Light Reading's inaugural Optical Networking Digital Symposium, which took place virtually in May.
The session featured presentations and panel participation from Telia Carrier vice president and chief evangelist Mattias Fridstrom and Huawei director of Optical Systems Competency Centre (Canada) Chris Janz. I served as moderator for the event.
Envisioning network automation, some people think only of fully autonomous, self-running networks that are devoid of human input and interactions – fully "closed-loop" operations. While this scenario falls within the automation spectrum, it is definitely at the extreme end of complexity. Technologies need to mature. Equally important, most network operators are simply not ready to hand over ultimate decision-making to machines. The essential point, though, is that there are many applications for automation along the spectrum that can be done today – and with big benefits for operators.
Use cases for optical networks
In his presentation, Huawei's Janz described three automation use cases in optical networks:
- Resource assurance: The telecom industry is moving to operations in which inventory, topologies and service maps are maintained and updated in real time by network software, replacing manual inventory management that is plodding and prone to errors. With accurate, real-time inventories, operators can plan for capacity expansion, including using AI software to aid in planning and predictions.
- Smart maintenance: This use case combines real-time data collection, machine learning, and predictive algorithms for operations, administration, & maintenance. Predictive health, for example, predicts degradations and faults before they result in SLA breaches and outages and uses machine learning and AI to recommend remedies (which, if desired, can be performed in closed-loop fashion). Another application is automated root cause analysis to reduce alarms and potentially eliminate erroneous fault tickets completely – a massive opex savings opportunity for network operators.
- Premium private lines: While much of optical automation focuses on operations savings, this use case is aimed directly at revenue generation. Here, Janz described using automation tools to "mass customize" optical layer services in order to provide differentiated services to customers, along with different (and thus premium) pricing tiers. The challenge, historically, is that managing mass-customized services is too complex to do manually, but software automation can do the job.
Lessons from COVID-19
In his keynote, Telia Carrier's Fridstrom gave examples of optical automation use cases in action – particularly around resource assurance and network health. Telia Carrier has been investing in network automation for several years and building its own automation software with a dedicated team. The global COVID-19 pandemic put those investments to the test.
Naturally, the network traffic impact from COVID-19 fallout has been massive. Mattias said that in three weeks, as countries shut down, the operator experienced 30% to 35% growth in traffic volume, as much growth as it would normally see for a whole year. This meant, in essence, that a year of capacity planning and provisioning was condensed into three weeks. Simulations were needed to understand impacts of new routes and automation helped the operator build capacity in the right place at the right time.
It was a lesson for those who still felt that the change from manual operations to automation was not needed. According to Mattias, planning and turning up capacity at such a rapid pace in response to COVID-19 impacts proved that optical layer automation is vital not just for the future, but also today.
The needle is moving
Lastly, through the presentations and panel discussion that ensued, the speakers emphasized that the role of automation extends beyond the broad use cases described here and is also an enabler in other major optical trends such as optical disaggregation, where it can play an important supporting role. As a specific example, Mattias cited the importance of open line systems in assuring vendor diversity and continuity of supply, particularly during global disruptions (such as disruptions caused by the COVID-19 pandemic). Open line systems with transponder interoperability allow operators to source transponders from different suppliers (with proper diversity planning). Software control and automation represent the glue that ties the terminals and the open line system together.
For open line systems and for optical automation broadly, both speakers agreed that there is still work to be done. But with real-world use cases to point to and accelerating events – such as the COVID-19 pandemic – the needle is moving firmly toward an automated optical future.
This blog is sponsored by Huawei.
— Sterling Perrin, Principal Analyst, Heavy Reading