Broadband infrastructure currently finds itself in a disruptive and innovative time, where past paradigms no longer hold true. During the pandemic, a lot of people have been depending on broadband, and living, working and playing from home has raised the bar on what we expect from it.
Unfortunately, the high user demand for digital services did not necessarily translate into higher profitability for telecommunication providers. This contradictory economic reality has kept the focus on efficiency gains and responding fast to changing conditions, which emphasized the relevance of network automation as a vital step in the digital transformation of an operator’s business.
SDN, cloud and AI/ML to boost network automation
To succeed with automation, a multidisciplinary approach is required to rigorously evaluate and continuously improve how broadband networks are built and run. Multi-generational, multi-technology networks have led to complex operational processes with increased data sources and volumes, heightening the need for comprehensive data management and automation strategies.
Software-defined networking (SDN) in access networkshas brought a step-change in automation and cloud capabilities. SDN takes network functions out of physical network assets, virtualizes and hosts them in the cloud, along with their associated data sets. This means operators can control network functions centrally and apply network automation for complex and time-consuming tasks while reducing manual interventions.
Also, SDN helps to create more self-aware, self-governing networks that can apply artificial intelligence and machine learning (AI/ML) to automate operations, improve service assurance, as well as provide detailed analysis for anomaly detection, troubleshooting and capacity planning.
A rising number of network tasks will benefit from data-driven optimization, especially thanks to the exponentially increasing amount of data available to train the algorithms on which AI/ML depends. The reliable capture, transmission, storage and processing of all these data sets is essential, as is the ability to stream and collect the vast data sets and make them available to operational support systems (OSS) and other applications via open application programming interfaces (API).
Single-pane-of-glass controller to simplify network operations
Automation of networks with nodes from multiple suppliers with different software implementations and technological support can be greatly simplified by using a single, common software-defined controller, such as those offered by Broadband Forum CloudCO-compliant implementations. The CloudCO brings the flexibility and programmability that traditional solutions are lacking.
What operators find attractive is the concept of a single-pane-of-glass SDN controllerwith open and standardized APIs. It’s very powerful as it allowsto plug in third-party modulesand manage entire multi-technology, multi-vendor networks from a single point.CloudCO compliancy makes it easy to connectto any type of access node: SDN-native fiber to the home/fiber to the distribution point (FTTH/FTTdP)nodes, disaggregated access nodes, white boxes,traditional simple network management protocol (SNMP)nodes, and it can even connect nodes without standardized interfaces by using proprietary software adapters.
Low-code automation tools to replace traditional scripting
For a programmable network withopen and standardized interfaces, a software-based automation engine is a logical next step.
Almost any task can be solved with the help of scripting. If you must perform several steps on different elements, however, the ease of use and ultimately, your daily efficiency will be reduced.
The power of integrated policy and workflow engines is in the ability to programconfiguration and maintenance tasks without writing a single line of code. They enable automation that can be repeatable and adaptable for a vast permutation of needs, services and processes. The low-code approach changes how operators go about writing, testing, executing and monitoring network automation.
Intent-based networking as the next step to full network automation
In policy-driven network automation, it’s important to collect feedback, observe network state and persist a configuration, even under conditions with frequent changes. Intent-based networking(IBN) is an emerging technology that aims to improve this by adding a feedback loop from the networkto keep configurations aligned over time,improve real-time visibility and reduce human error.
Intent-based automation helps to define the outcome operators want to achieve in terms of a service definition or service-level agreement (SLA), and the network will translate the business intent into policies that can be automated and applied consistently across the network. Without the need for the OSS to micro-manage everything, the network can be more intuitive about resolving requests, translating a service requirement into an implementation.
An intent-based tool is much more effective as all actions are declarative and can consider that the realization can change over time. The network will also self-monitor and auto-adjust when it finds itself in a sub-optimal state compared to the intent.
Use case prioritization to drive implementations of SDN
Automation helps in every stage of the network lifecycle, and each stage holds unique opportunities and use cases for optimization.
First of all, customization to the network environment is key, such as programmable service modeling, WebUI programmability, cloud-native application integration, third-party device modeling and flexible cloud deployment model. Intent-based models can be easily programmed, configured and kept up to date. This uniformity of SDN greatly simplifies network operations. Once operators start to define and collect telemetry, metrics, health, KPIs and alarms, they can close the loop to implement advanced use cases.
Examples are all kinds of moves, adds and changes in the network, including card replacements, node swaps, customer activation of a new optical network terminal (ONT), fiber moves due to infrastructure changes, subscriber upgrades; technology migrations from gigabit passive optical networks (GPON) to XGS-PON and updates of existing services and profiles.
But service assurance, network health and customer experience can also be automatically analyzed, optimized and restored, such as mobile cell site alignment for fixed wireless access, PON bandwidth optimization, pro-active subscriber campaigns to improve customer experience, anomaly detection based on data-driven machine learning, and root-cause analysis and troubleshooting diagnostics to help fault resolution.
The automation of these processes represents a major opportunity for operators.
A step-by-step evolution is key for the adoption of SDN
The move to SDN and the cloud is an access network transformation that happens once every 20 years. When, what and how to implement virtualization and IT/cloud principles are challenges that deserve up-front attention to make the resulting achievements viable in the marketplace.
In most cases, the journey to SDN and the cloud will be a long-game transformation. Operators will incrementally add SDN-enabled elements and use cases over time alongside their trusty, valuable legacy assets and operations, creating a hybrid traditional-virtual environment.
The strengths of any cloud and SDN offering can be evaluated based on their system integration and software development capabilities. Building a customized SDN solution requires disciplined systems integration to keep together all the elements in a working solution.
That’s why there’s a vendor-agnostic system integration and software development approach where the operator IT environment and all network nodes—third-party, virtual and legacy—find their place. This avoids ending up with isolated islands of SDN/network function virtualization (NFV) deployments, long integration cycles or forklifting the whole system.
Operators can apply SDN where and when it provides value and pace their investment to “grow-as-they-learn” with incremental deployment approaches that help achieve their goal of achieving high-performance network automation.
