How Can Data Enhance Coverage Optimization and Extension for Mobile Network Operators?

As mobile network operator, you have a wealth of data available to help optimize and extend your network coverage and connectivity. The problem is that, with the current approach, the data is scattered in different silos. Unless we can bring this data together, you won’t be able to obtain the insights you need to do things better and to truly leverage it to enhance the coverage we deliver. And, you won’t be able to unlock valuable new use cases.

The Growing Challenge of Network Optimization

Optimizing mobile network coverage has become an increasingly challenging task. Networks have become more complicated, with new services to support. Maintaining connectivity with the right service levels, even as you add new capabilities is difficult. Delivering new capacity is something that must be got right, so you can ensure your investments are a success.

So, you have to continuously optimize performance for the areas already covered, ensure that conditions are optimal for every service launched or enabled for others, and extend coverage to new locations. 

Get Smarter About Coverage Optimization

Currently, most MNOs use static data to optimize coverage, such as population numbers and density, local topology, signal strength demands, and so on, while leveraging network data to track performance and experience as part of assurance programs.

However, that approach is limiting. Yes, it gives the ability to tune around baselines and to troubleshoot when problems emerge. But it depends on isolated data sets, and this means you cannot leverage a unified data set that can drive deeper and richer insights: MNOs need the ability to get a lot smarter about how they manage critical mobile network coverage.

Augment Existing Data Set with Others

To achieve this aim, a new approach to data is required. You need to be able to leverage the other data sources that are also available. These include information from Performance Management outputs – generated by systems in the OSS, which include telemetry data, status indicators, and much more. There is also Configuration Management data, Fault Management data, information from ticketing systems and so on.

Each data set provides a different perspective. Traditionally, they have been viewed separately, essentially in silos. As a result, information that can complement other data cannot not easily be correlated.

Yes, you can learn about network performance from monitoring data, while also checking system status from alarm handling solutions – but you cannot view these essential inputs together.

In turn, that means that you can’t use all the available information sources efficiently to plan and optimize the best coverage for a given cell. There is information regarding traffic patterns; there is also information regarding service quality, RAN performance and so on – but the desired unified view is absent.

Combine Data Sets into a Unified View

But combining node level performance data (PM) with probe data like subscriber experience network health statistics, together with alarms (FM), infrastructure metrics and logs offers deeper, actionable insights.

For example, with information such as:

• Customer experience indicators
• Geolocation data
• RAN signaling metrics
• System availability through time
• System alarms
• Cell usage and activity
• Service performance indicators for different applications

You can create a more accurate picture of how services in a cell are experienced, the delivery and demand over time – and discover whether there are opportunities for enhancement.

And you can also ask — and answer — questions like:

• Is there sufficient capacity to meet peak demand when users consume the most intensive services?
• How do different services perform?
• Are we meeting the right KPIs across all system variables?
• Do we need to enhance capacity to meet increased demand?
• Is the site capable of supporting new services with different KPIs and QoS requirements?

So, with this unified view, you can make informed decisions that are based on analysis of all available data, through a single pane of glass – and also use this enhanced dataset to drive accurate automation.

With that in mind, let’s consider how this data can fuel new use cases and optimization tactics.

Optimizing Event Coverage

Temporary coverage for annual or infrequent events needs to be delivered accurately and efficiently, balancing both costs and the experiences delivered. Visitors to the event have high expectations, so there’s both a need to meet those and to improve outcomes for the next edition.

Armed with this newly combined dataset, you can use PM data and network performance KPIs to really understand both how the network delivers services as experienced by subscribers and how the functions responsible for delivering those services are performing. Geolocation data can also be used to track service demands and user flows.

In effect, you can take a holistic view, as well as to isolate different domains. As a result, you can see, in real-time, how people move around the site, where they pause or dwell (to watch those bands on the main or satellite stages) and what they are doing during their visit.

During the event, this enables you to troubleshoot and manage connectivity – but the captured data also allows you to plan more effectively for the next. Did everything perform as expected? What happens if audience size grows (or shrinks). You can use the data to model different scenarios, based on real audience behavior and how people and devices accessed the network – which means you can design the best coverage for the next event, based on lessons gained from the previous edition.

RAN Lifecycle Management

Making the right investment choices is always a challenge. If users complain about coverage in a particular cell site, does that mean there is insufficient capacity, or that something isn’t working or configured correctly? In the former case, money is required to add new transceivers – but in the latter, adjustments can tune existing resources, pushing expenditure further down the line, or even avoiding it altogether.

If you have a truly aggregated dataset, viewed through a single pane of glass, you can make these decisions more accurately. Is the available spectrum being used correctly, or do cell settings need to be changed? Do power cycles actually match demand? Are all users affected, or only those with certain devices?

It is only with easy to access data that you can truly explore the real situation. You can also use historic data to explore performance under different scenarios and note any changes that have occurred through time.

As a result, you can more efficiently and accurately optimize our resources, based on all relevant data, tuning coverage so that desired outcomes and experiences are delivered, intelligently.

Towards Accurate, Continuous Optimization

And, of course, this takes us back to the original question: How can data enhance coverage optimization and extension for mobile network operators? With the answers to the questions above — and more — you can review site performance, and both implement optimization strategies and also plan how to extend and enhance coverage for each site in your networks.

This can all be achieved, provided we shift to a ‘data-first, insight-driven’ approach. That means bringing these different sources of data together – using advanced DataOps processing, so that they are ready for analysis and interrogation – so that you can automatically generate valuable insights that go far beyond what you can do with the legacy, siloed approach.