Performance Management (PM) data provides a constant stream of information that helps MNOs to optimize their networks and to evaluate their behavior. By collecting counters from different nodes, you can gain deep insights into the overall experiences delivered.
However, you cannot correlate this information with the experience of actual subscribers. For this, information captured by network probes is required, so that user data, such as MSISDNs can be obtained, enabling per-session analytics.
One MNO — a leading provider of mobile services with more than 20 million subscribers - recognized that maintaining PM and probe-based systems as silos was creating friction that impeded troubleshooting, slowed network optimization efforts, and artificially increased costs. It sought a new approach that would consolidate these separate platforms into a single, integrated solution.
How was this achieved and what outcomes have been delivered?
Problems with Data Management and Correlation
The problem is not only that the solutions existed as silos. They also need to serve the interests of different teams and users. The RAN team has one perspective, the customer care another – while the IMS and core teams have their own requirements. Meeting all of these needs is demanding – but it is also essential to deliver the end-to-end performance on which the MNO’s reputation as a market leader depends.
In addition, like most MNOs, the network spans multiple generations of technology and continues to evolve as 5G SA and advanced enterprise applications with specialized performance requirements are rolled out.
This means that the landscape is not static. New PM counters are regularly provided by vendors, adding to the thousands already available, and these are typically formatted in proprietary or vendor specific schemes. At the same time, network protocols and interfaces continue to change with each mobile generation. In this case, the MNO had a PM solution from one of its primary network infrastructure vendors, while depending on Polystar for its network probe and assurance platform.
Of course, data collected by network probes has its own formats, according to 3GPP and other specifications. As a result, while there is a host of data available from these different sources, it exists in different sets.
In addition, the different teams had been accustomed to making sense of these, relying on their accumulated expertise – but if data could be presented in other ways, the user audience could be extended, enabling more people to contribute to optimization activities.
To obtain the new, comprehensive picture the MNO sought and to enable new ways of exploring data and securing insights, these data sources would need to be combined, correlated and then presented to system users.
Solution Design
– Supporting Existing Work Practices while Unlocking new Perspectives and Insights
The key to a new, consolidated system that spans different sources of data are efficient, effective data pipelines. In turn, this depends on the ability to ingest, process and deliver data to an appropriate data warehouse, ready for interrogation and analysis. In other words, DataOps is a necessary requirement.
DataOps provides a complete methodology for the processing of different data sources and formats, delivering normalized pipelines, ready for analysis. Polystar’s Network Assurance solution was already deployed by the MNO to provide the crucial network probe data - and now also includes powerful DataOps processing capabilities.
As such, extending this function to include PM data by the addition of the Performance Management capabilities was the key step to assure the operator that a single, unified solution could be realized. And, when combined with PM data for the RAN, the end-to-end vision sprang into life.
The needs of different stakeholders had also to be considered as part of the implementation. While the RAN team with its deep levels of experience and knowledge expected to be able to review large volumes of detailed counters and KPIs, other users wanted tools that enabled them to move between higher-level information, all the way to drill-down, with prompts to guide their data queries and visualizations.
Finally, the MNO also had an eye on automation. The combined platform also had to support this vision, with the ability to support triggers that could initiate actions autonomously, driving closed-loop operations and assurance.
Outcomes and Delivery
KALIX provided a solution to the need for consolidation. With the integrated DataOps capabilities, it combines the ability to process probe and PM data in a single, integrated platform – and other data sources could, in the future, also be added. When data is ingested, it is normalized and made available for analytics functions, as well as formatted into pipelines for other activities.
Because a pipeline catalog is included, the integration of different data sources is facilitated. What’s more, when a key vendor changes counters included in the PM data their solutions generate, the combined Network Assurance and Performance Management solution - KALIX - automatically adapts to the new information source.
KALIX also supports different portals and views, which provides the flexibility to enable those users that require access to deep-level information to zoom to that view, while providing other users with different navigation paths and visualizations.
The Result
- First, the objectives of the consolidation have been realized, and the legacy solutions can be retired, with all PM data now processed in the expanded Polystar solution. The end-to-end view and combination of probe and PM data enables users to correlate system performance with user experience, leading to better outcomes for subscribers and faster problem identification and resolution.
- Second, a reduction in operational costs has been achieved, as a single solution provides the required functionality, reducing the overall number of systems that must be supported.
- Third, the needs of different stakeholders have been satisfied because different views and levels of analytics are integral to the user experiences offered – which accelerates problem resolution, delivering further time saving, enhanced insights of customer experience, and cost reduction benefits.
However, in addition to these considerations, the workflows and outputs enable automation of key tasks, such as Root Cause Analysis – boosting efforts towards automation for service assurance.
So, the ambitions of the MNO have been met. Polystar’s solution, supporting both probe and PM-based analytics and processing provides the single platform desired – and the foundation of widespread analytics and automation functions, across key processes and for nearly 500 users in the organization.
