Data-Driven Transformation in Telco Operators | White Paper | Polystar
The Telecom Industry is facing a deep transformation. Read about the CSP status, as well as the drivers for transformation. Download here
Why is automation of service assurance and data management is the foundation of autonomous operations? Learn more here
Network and operational automation have long been a goal for telcos. Automation can reduce costs, enhance agility and reduce dependency on manual intervention – all measurable targets that can be identified and tracked with performance metrics. Different operators have pursued different approaches, with the result that, while some are more advanced than others, most have already secured some benefits.
However, automation has now become a strategic target for telcos, with many following the path suggested by the TMF in its six-step model for autonomous networks. While earlier efforts had been more ad hoc, mounting pressure on telcos has made automation essential.
That’s because, with recent investments in 5G (both 5G NSA and 5G SA) to monetize, and the combination of increased data volume generated by network processes together with the dynamic operational performance new service require, manual intervention can no longer deliver the service levels and customer experiences demanded.
Efficient automation depends on data - of sufficient quality, quantity and variety - so a key element of this transformation journey is to effectively secure sources of consistent data to fuel automation efforts.
One key source of such data is the same information that has, historically, empowered real-time network and service analytics. This consists of information captured from the Control and User Planes of fixed and mobile networks, as well as Performance Management and Configuration data, which is obtained from feeds from systems and entities in the OSS layer.
Together, these data provide information that allows sessions and events to be correlated with both individual users and devices, as well as pinpointed to specific operational entities in different network domains – access, transport and core.
But this data, from those different sources, needs to be effectively managed and integrated so that it can be accessed and put to work. That demands an entirely new approach, bringing advances from DataOps processing to bear. With DataOps, data from multiple different sources can be consolidated and homogenized so that it can be made accessible to other tools and processes.
Armed with this normalized data set, telcos have been able to build upon service assurance programs, accelerating root cause discovery for problems that affect users, services and network components, and enabling them to proactively manage the experiences they deliver. But this data offers much more – it can also enable automated responses to problems, so that they can be resolved without manual intervention.
Automation depends on awareness of what’s happening and the possible outcomes and responses that can result. This is applicable at all levels – so, if a cell site is experiencing congestion, specific alarms from the RAN will provide notification of the current status – and can be used to trigger an increase in capacity in response. Whereas previously that had been a manual activity, automation algorithms can be developed and deployed, such that the appropriate reaction to the reported events can be applied.
The same principles can be applied to a wide range of events and issues that are indicated by the data that is captured for service assurance.
So, given its importance, what do telcos need to secure this valuable source of data?
Essentially, we need to obtain analytics information from every domain in the network. Without that, we can’t assemble an end-to-end view or be sure that we are covering all elements and data that are associated with a specific service or event.
Next, we must combine network and service context effectively, so that we can track and trace events to their impacts. For example, a user may experience a problem with a specific service – call drops, or perhaps they are unable to stream content effectively. Is this a problem with the network (the infrastructure), the service (the content delivery or the VoLTE session), or the device that is trying to access the service?
Recognizing the need for truly integrated data, from multiple sources, we’ve also incorporated Performance Management data from the OSS layer, thanks to our adoption of DataOps processing that enables data ingestion from multiple sources. With that, we can use information from operational systems directly — status, configuration, capacity, and much more — to expose the full range of inputs that help us to determine and isolate problems and their solutions.
So, in the examples above, the problem could reside in the systems, the service or the network, or just one of these – and reports from different elements may conflict (the network may be operating as expected, but actually there’s an issue with the device in question). By combining these sources thanks to DataOps processing, we can eliminate possibilities until we settle on the correct cause of the problem.
Third, we need to have vendor independence, because many operators are adopting solutions from different providers, creating diversity in their supply chains and reducing dependencies on single suppliers.
With all of these elements in place, we can start to think about implementing automation in practice. As we noted earlier, automation enables responses to detected events to be triggered without human intervention. So, by parsing all the consolidated data — from the domains, entities and devices — responses can be initiated from the range of possible outcomes.
Of course, there are many such options available, so the correct one needs to be found and triggered, once all available information has been processed. With the problem identified, and action taken (more capacity), appropriate subsequent actions can also be initiated (once the congestion has been relieved, are there now fewer users in the cell, so that power can be reduced to previous levels, for example).
The enriched analytics data, then enables both new analytics capabilities as well as automated assurance for operators. At Polystar, we’ve gone further, implementing multiple automation use cases that help operators to meet this challenge head-on.
We’ve accomplished this, because we leverage more than 30 years of telco analytics leadership - together with the practical experience of implementing automations by our parent company, Elisa.
This sets us apart. Every automation we make available has already been proven in practice: because Elisa uses its Tier 1 network as a living laboratory, developing automations that can be proven in the real world, we can then bring them to operators via the global presence of Elisa Polystar, and delivered via a common, integrated analytics platform. So, we convert theory into practical solutions that are deployed in our operational network.
As a result, our automated multi-vendor assurance solutions support comprehensive FCAPS (fault, configuration, accounting, performance and security) functionality, as well as advanced monitoring and management to support root-cause analysis, network planning and design tasks.
We’ve also accelerated investment into AI and ML, implementing algorithms that unlock new, advanced automation use cases, based on predictive analytics and the discovery of contextual insights for optimizing real-time operations.
The result? An integrated solution spanning all generations of network technology, together with all major data capture methods that also includes a range of pre-packaged automation templates and use cases. It provides an open platform approach, so that operators can implement their own automations through API integration – fueled by the data ingested by our unified multi-domain telco analytics platform.
Proven in multiple reference customers globally and Elisa’s own networks, our analytics solutions allow users to customize reporting experience to uniquely match their needs. With the exceptional performance delivered, operators can fully monetize and optimize their investments in networks and services, while ensuring consistent customer experience.
We provide cloud-native, AI-driven solutions to mobile operators worldwide. Our innovative software transforms telecom data into smarter decisions and actionable insights.