RAN Resources & Insight

For the better part of the past decade, the governing logic of RAN planning was simple:
Mobile data traffic was growing at 40–50% annually, and investment strategy was
driven overwhelmingly by capacity. That era is over. Traffic growth has moderated
globally to approximately 20% per year and is expected to continue stabilizing. The relief on capital expenditure is real, but it is also a trap.

Operators that interpret slowing traffic growth as a signal to reduce strategic ambition will find themselves unprepared for three simultaneous and transformative forces:
1. The shift in focus from capacity to revenue
2. The AI-driven progress from downlink-centric to uplink-sensitive traffic patterns
3. A multi-year technology migration from legacy 2G/3G to 5G and eventually to 6G from the 2030s onward.

This paper examines each of these forces and offers a framework for how Operators can reorient RAN planning from a cost-efficiency discipline into a revenue-generation capability, while protecting the network performance and customer trust that is the ultimate foundation of long-term competitiveness. FL Omnitele stands ready to support Mobile Network Operators (MNOs) through this transition.

For much of the past decade, the standard response to degraded mobile network performance has been to build more capacity. When user speeds dropped or connectivity deteriorated, the industry's default response was to add more spectrum or commission additional sites. In the era of 5G-Advanced and tighter capital constraints, however, this approach is no longer financially sustainable or strategically sound.

Modern Radio Access Networks have reached a level of multi-layer, multi-band complexity that traditional monitoring tools can no longer effectively manage in operational environments. As a result, operators often find themselves investing heavily in additional capacity while actual user throughput remains stagnant or even declines.

This paper examines the shift from reactive monitoring to precision diagnostics. By using artificial intelligence to decompose the user experience into its underlying technical drivers, operators can identify significant performance improvements without additional capital expenditure. We demonstrate that moving from broad network averages to site-specific root cause analysis can unlock substantial gains in user throughput while maximizing the value ofexisting network investments.

Omnitele conducted an independent benchmarking campaign in Albania in June 2025, assessing the network quality and service experience. The benchmark measurements compared the voice, data and application services in ONE and Vodafone networks. Measured areas included 10 cities and main connecting roads between the cities.

Both operators have started large-scale 5G deployment in Albania. 5G deployment has had a significant impact on the service experience. Typically, 5G deployment directly impacts the download speed benchmark, as 5G can provide a significant boost to the download performance, whereas the impact on upload speeds is somewhat more limited.

Between the Omnitele benchmark in 2024 and 2025, the average download speed has improved by over 70 %. Also, average upload speed has improved by over 30 %. Application performance results have also improved, allowing the users to load web sites faster, have faster access to higher quality video streaming and post content, such as pictures, quicker to social media.

FL Omnitele conducted an independent assessment of network quality and service experience benchmarking campaign in Albania in February 2024. The benchmark measurements compared voice, data, and application services across the ONE and Vodafone networks. Measured areas included 11 cities and the main connecting roads between them.

Evaluation of end-user perceived service experience revealed that ONE Albania had the best overall service experience in the benchmarking campaign. ONE scored 95.25 points (out of 100), whereas Vodafone scored 91 points. ONE outscored Vodafone in each service experience category in the measured areas. Read the full report from