Tackling the Burden of Poor Data Quality with Tagging

By: Jan Ulc, CROSS Network Intelligence

The 5G era brings transformation towards a new, service-oriented architecture for operators globally. The new network architecture is designed to enable the support of an increasingly wide range of services, with an equally wide range of different performance requirements. It’s also supposed to introduce new dynamic capabilities, so that services can be composed, delivered and orchestrated on the fly, before being shut down or recycled. And, lest we forget, 5G isn’t just for mobile: this new architecture will become the foundation of fixed as well as cellular networks.

It’s a big vision for operators. To achieve this level of agility, they will have to transform their operations and processes. Simply acquiring the ability to spin up new services dynamically won’t be enough, operators will have to be able to ensure that they deliver them, at the right time and place, and with the right experience required.

This presents challenges because customer experience is directly linked to service quality and performance. In turn, this requires careful orchestration of the resources on which services depend. Operators must know what resources they have available, where they are and how they are being used at any given moment. These resources include services, transport infrastructure, physical assets, information regarding their location and routes such as wayleaves, and metadata such as street addresses, as well as virtual elements that may be required when a service is composed

And, they not only need to know this information, but they also need to be able to obtain it, when required and in real-time, so it can be shared with appropriate business processes. That’s because a key goal for the next-generation network is to enable automation. In this context, automation means that a service could be ordered by a customer, with the necessary actions taken to deliver and manage it being enabled automatically with zero-touch intervention.

It’s not just customer orders, however. A series of conditions or events could also trigger the establishment of a new service or modification and extension to a live session, dynamically. Again, the orchestration and service platform must be able to check what elements and resources are required to fulfill these demands, whether they are available, and then to instantiate the change by composing the necessary elements in the right way. Operators need a dynamically updated, real-time view of resources, available as a complete picture: a single source of truth for their networks.

To understand and to obtain this complete picture requires accurate, high-quality data that is accessible to all business systems, as well as a consolidated record of all such assets. This must all be consolidated into a single data model—in other words, a comprehensive network inventory system. 

Data problems

The problem with this is that few operators have access to such data. Worse, few also have access to the single, consolidated inventory of their network assets that is required. Instead, many have multiple, disconnected inventory systems, populated with incomplete and poor-quality data. As a result, plans to deliver on the 5G promise and dynamic service orchestration are likely to be undermined. 

The first step is to recognize that the problem exists. It turns out that failure to recognize this problem is surprisingly, if unintentionally, common. Many operators have absorbed legacy practices and systems, making do with different silos of data, even as they grow their network and subscriber base. This situation often results from mergers and acquisitions. Because until recently many operators had manual processes for activating and delivering services, the problems of such approaches were not apparent.

And, even if the difficulties that such legacy practices have caused are known and understood, the task of rectifying them can often seem too huge to tackle in the face of other network evolution and transformation initiatives. This is the well-known problem of inertia: we understand the problem but cannot take action, because we fear the consequences or think that it will demand too much effort.

Finally, data may well be available, but it may be scattered across different silos, in different databases and may also be incomplete and imperfect. Some operators have paper-based records or an array of spreadsheets, so the data available exists only in different, incompatible formats. Making sense of this can be time-consuming.

The data needs to be brought together, but the imperfections also create a significant obstacle. What’s needed is a means of processing this data, regardless of any imperfections, so that operators can embark on the task of transformation and inventory data consolidation but without the need for a seismic approach.


Latest Updates

Subscribe to our YouTube Channel