• Topics
  • Research & Analysis
  • Features & Opinion
  • Webinars & Podcasts
  • Videos
  • Dtw

How to enable smarter investments for greener network rollouts

The AI Smart CapEx for green and efficient network investments Catalyst is using AI to automate and optimize the rollout of radio access and fiber networks and unlock major cost savings

Alasdair Riggs
11 Jun 2024
How to enable smarter investments for greener network rollouts

How to enable smarter investments for greener network rollouts

Commercial context

As capital-intensive businesses, CSPs are always looking for ways to optimize their investments in infrastructure – and as such need to perfect their understanding over time of both the capital costs (capex) and operating costs (opex) of their network rollout plans. That understanding helps them manage energy costs, environmental pressures, the role of tower companies with shared infrastructure, as well as the decommissioning of old equipment, among many other challenges.

Forecasting network traffic can be difficult, though, particularly as most planning tools only account for the CSP’s internal datasets using capacity models provided by their infrastructure vendors. Furthermore, with marketing staff focused on customers (and increasingly devices), their forecasts and projections are hard to map onto network traffic load and properly align with the network team's plan of action.

In light of these challenges, most CSPs have embraced the ‘golden sites’ methodology that focuses on the network elements as revenue-generating components. This approach, however, falls short of capturing customers’ sentiments and shifting usage patterns, which drive uneven demands on the CSP’s network resources. This is a major issue given that the customer experience across an operator’s coverage area plays a significant role in the operator’s ability to retain high value customers.

The solution

This Catalyst aims to improve the capital allocation process by forecasting traffic more accurately, and using AI to automate and optimize the rollout of radio access networks and fiber. The solution has built a holistic traffic forecasting model by using a CSP’s internal datasets, such as CDRs (call detail records), network and user equipment KPIs and CRM data, along with external datasets, such as third-party information tracking the user experience, people’s movement patterns, buildings and terrain.

The first step was to provide the CSP’s marketing team with a user-friendly tool that will enable them to understand technology adoption trends in the customer base, and systematically forecast uptake of new use cases for which there isn’t proper historical data. This self-service solution also enables decision makers to perform what-if scenarios across different time horizons in annual, semi-annual, quarterly, or even monthly reviews.

“We can use the dashboard to zoom in to the cell level pinpointing exactly where demand is surging,” explains Fernando Carrasco, CEO at Locatium.AI. “Our approach considers the full potential of 5G beyond traditional use cases. We incorporate advanced applications, such as virtual reality gaming, autonomous vehicles and FWA (fixed wireless access) which are expected to significantly drive up data usage as these technologies become more prevalent.”

The next step was to translate high-level commercial projections into street level traffic load, while also considering how the competitive dynamics in the market will impact future traffic levels. To that end, the solution accounts for differences in the network customer experience across operators

As this analysis involves a huge increase in computational complexity (in the range of tens of thousands of central processing cores), it requires an automated cloud-based system. To measure the effectiveness of the solution, the project is tracking deviations between the solution’s traffic forecasts and the actual traffic demand per site and technology over a period of time.

Application and wider value

The Catalyst has piloted the solution with an operator in Brazil with more than 5 million subscribers, the results of which suggest the solution could reduce CSPs’ capex and opex by 15-20%, while catering for traffic growth and maintaining a good customer experience. For a medium-sized CSP, capex efficiency gains of 15-25% on new network investments would translate into savings of hundreds of millions of US dollars a year. As well as reducing opex (including electricity consumption) on the new infrastructure by 15-20%, the solution could achieve significant spectrum and electricity savings on the existing infrastructure, thereby contributing significantly to CSPs’ efforts to reduce their greenhouse gas emissions.

If applied across the global telecoms industry, the solution could generate capex savings of approximately US$32.4 billion annually and about US$11.2 billion in energy savings annually. The step change in energy efficiency would also support global climate goals and position the telecoms sector as a leader in sustainable practices, potentially setting a standard for other industries.

To learn more about how this initiative could optimize autonomous network rollouts for radio access and fiber, and help CSPs reduce their carbon emissions in the process, please click here to view the project space on the TM Forum website.

AI Smart CapEx for green and efficient network investments