The convergence of terrestrial and non-terrestrial networks is no longer optional; it is a strategic imperative on the path to truly ubiquitous connectivity. But the more we rely on satellites, the more demands we place on them. Beam management is one area slated for change, with many satellite companies actively planning upgrades. This paper outlines a new approach to beam-hopping that will enable satellites to serve multiple areas fairly and efficiently. 

When 3GPP Release-17 defined the specifications for satellite-based connectivity as part of 3GPP specifications, it enabled the path to integrated architecture for NB-IOT/5G NR and Non-Terrestrial Networks (NTN). Capgemini took up the challenge and started to explore the evolution of satellite networks based on 3GPP specifications. Our innovation accelerates the business case by unlocking new use cases in IoT, mobility, logistics, and rural broadband – where connectivity gaps traditionally hinder adoption. For satellite companies that need ever greater efficiency from their fleets, it could be a significant step forward. 

The industry challenge: Capacity from space

Non‑Terrestrial Networks include satellite constellations in Geostationary Earth Orbit (GEO), Medium Earth Orbit (MEO) and Low Earth Orbit (LEO). LEO satellites offer advantages such as reduced latency, better link-budget and path-loss compared to GEO and MEO satellites, and are thus increasingly seen as the key to extending 5G beyond traditional coverage boundaries. Multiple LEO constellations are currently being planned, with use cases such as: 

  • Maritime and aviation connectivity   
  • Rural and underserved areas  
  • Public safety and disaster management   
  • Remote industrial sites (energy, mining, logistics)   

As the industry moves toward truly ubiquitous connectivity, LEO satellite systems are becoming a core pillar of 5G evolution, and are even laying a foundation for 6G. However, delivering coverage from space also introduces unique challenges. 

The need for beam management 

How do you get maximum coverage from each satellite? First, satellites today deploy multiple beams – you can picture a honeycomb pattern on the ground, with each cell corresponding to one beam. But modern 5G satellites take this one step further. In order to optimize satellite payload weight and limit the amount of energy consumption, the number of beams deployed on a satellite are typically much smaller than number of cells covered by it, meaning that one beam is responsible for multiple cells. 

This strategy requires an elaborate scheduling framework that dynamically maps cells to beams, such that a particular beam illuminates multiple cell clusters on a time-sharing basis. Furthermore, the traffic load across cells may fluctuate, and beam allocation must adapt to these dynamic changes. 

Figure 1: Overview of Beam-hopping in NTN 

Satellite payloads must remain lightweight and power-efficient, even as they serve thousands of dynamically changing ground cells. This makes intelligent beam management essential: satellite beams must be allocated, adapted, and optimized in real time to match varying traffic patterns and ensure seamless service continuity. 

Capgemini’s experience in beam management

At Capgemini, we have been working to make this a reality within the 3GPP 5G standards and the O‑RAN ecosystem. Our recent IEEE paper, “Beam Management for 5G FR1 LEO-based Non-Terrestrial Networks” at the IEEE Future Networks World Forum (FNWF) 2025, addresses one of the most critical challenges in satellite‑based 5G: how to intelligently manage satellite beams to deliver capacity where and when it is needed. 

Our paper presents a novel beam-hopping and resource allocation framework designed specifically for 5G NTN. Using an O-RAN-based architecture and an intelligent Beam Allocation Module (BAM), the approach dynamically assigns satellite beams to ground cells while complying with 3GPP standards. Simulation results demonstrate how this technique significantly improves capacity utilization and enables more efficient use of spectrum and satellite resources.  

In practice, operators face three intertwined challenges, as shown in Table 1: 

Challenge Description 
Limited satellite payload and beam resources A single LEO satellite must cover a very large footprint on the ground. To keep SWaP (Size, Weight and Power) under control: 
– The number of satellite beams and Payload Total Power (dBW) is fixed 
– The number of ground cells that need to be served is much higher than the number of beams 
This means a beam must be shared in time across multiple cells – a concept known as beam hopping. 
5G NR scheduling and timing constraints Beam hopping needs comply to 5G NR standards such as: 
– Transmission of mandatory signals such as Synchronization Signal Blocks (SSB), System Information Blocks (SIBs) and Random Access (PRACH)    
– Transmission of Control channel (PDCCH), Data channel (PDSCH) and HARQ acknowledgements  
Each cell must be illuminated by a satellite beam at those specific time instants. If different cells sharing the same beam need mandatory signalling at the same time, the system risks: 
– Coverage holes (user can’t find or access the cell)   
– Protocol violations (non‑compliant behaviour vs 3GPP spec)   
– Unpredictable performance 
Highly dynamic and uneven traffic NTN networks cover vast geographical areas and are characterized by: 
– Uneven traffic demand: some cells could be highly loaded while others may not have any traffic 
– Dynamic traffic variations: high demand for a particular period followed by long periods of inactivity
Static or semi-static beam allocation strategies fail to adapt to demand and waste precious satellite resources, leading to: 
– Underutilized satellite spectrum and payload   
– Poor user experience during peak demand   
– Difficulty in meeting SLAs for enterprise customers 

Table1: Challenges in NTN networks

In summary, there is a need to dynamically share satellite beams across many cells, comply all 5G scheduling and timing constraints and maximize capacity based on traffic demand. 

In part 2 of this blog, we’ll demonstrate how we did just that.