Non-terrestrial networks (NTN) are wireless communication systems that are operated above the Earth’s surface and involve satellites at low Earth orbit (LEO), medium Earth orbit (MEO) and geostationary orbit (GEO), high-altitude platforms (HAPS) and drones. All of these components are deemed essential to providing seamless coverage and bringing coverage even to remote areas.
Recent research from the Global mobile Suppliers Association (GSA) has shown that 50 operators in 37 countries and territories are planning satellite services, with 10 operators having already launched commercially.
Many are describing this as a new space race among tech companies who are looking to deploy satellite constellations to deliver high-speed internet services to emerging markets and business customers.
Each of these companies has recognized the potential of these satellite constellations to not only provide Internet connectivity to rural areas but satisfy the global networking services of tomorrow.
Historically, the satellite industry has spent significant sums on the backend to incorporate new satellite technology within prevailing telecommunication standards and practices but, today, 5G provides a networking architecture from the get-go.
However, when it comes to deploying NTNs, there are several unique challenges not faced when deploying terrestrial networks. These range from those which must be taken into consideration for their impacts on network performance, such as speed and orbit altitude and attitude, to the more mundane, such as lack of access for maintenance or to fix faults.
Traditional Geostationary Orbit (GEO) satellites orbit at an altitude of approximately 35,000 kilometres above the Earth's surface, and while Low Earth Orbit (LEO) satellites are much closer, their orbits still typically range from 160 to 1600 kilometres in altitude, traveling at around ten thousand miles per hour.
Dealing with these challenges is critical to the success of delivering satellite 5G services, as otherwise it will be impossible to provide the necessary network performance on which these offerings will succeed or fail. If satellite providers fail to achieve the required latency, coverage and throughput to match the expected service requirements it will be impossible to satisfy user experience demands and therefore risks making the business case unsustainable.
Therefore, it goes without saying that to integrate satellites into 5G networks, operators need the right technology. This starts with selecting the right antenna technology, radio front end, and a physical layer processor capable of ensuring a high-reliability link. This is particularly important in satellites where channel capacity is highly constrained in comparison to terrestrial networks.
The Architecture Question
Since Apple's initial announcement of emergency satellite services using Globalstar's GEO network in September 2022, the industry has seen a number of significant developments.
SES has acquired Intelsat, Lockheed Martin plans to launch 5G Low Earth Orbit (LEO) satellites later in 2024, and Starlink continues to expand its coverage and user base. However, as 5G LEO satellites look set to enter the mainstream, a crucial question remains for the industry: how to best architect these networks.
There are essentially two architecture options: transparent (also known as 'Bent Pipe') or regenerative. How the industry moves forward in this matter will significantly affect the performance and success of 5G satellite services for the next decade and beyond.
The main factors driving the decision between Regenerative and Transparent architectures include:
Coverage – where the users are located, including remote, austere and oceanic
Bandwidth – Optimising link capacity to get the maximum bandwidth for users
Latency – The delay and responsiveness characteristics of the network, and the sensitivity to this latency of users and applications
Link performance – The overall user experience and how this is impacted by signal-to-noise ratio and power
To effectively navigate the choice between the architectures, it is useful to understand the differences between the two approaches in more detail, including the benefits and restrictions that they offer.
Regenerative vs Transparent
Essentially, the difference between regenerative and transparent architectures relates to where the 5G base station functionality is implemented. Under the Transparent model, the base station is in the gateway on the ground, with the satellite acting as a relay. The means that two over-the-air “hops” are required for every communication between base and user equipment (UE). For the Regenerative model, all base station data processing is performed on the satellite, so only one “hop” is required.
Both approaches have positives and negatives. With transparent, the main benefit lies in having lower demands on satellite payload processing, meaning lower power and cheaper satellites, although it is estimated that this cost differential for the payload is only around 10%.
For the Regenerative approach, there are a range of benefits enabled by placing the base station and additional compute power on the satellite itself. These include native support for inter-satellite links (ISLs), which significantly enhances coverage as it allows the satellites to still maintain communications when they are out of direct contact with ground stations. As Regenerative systems can perform signal regeneration, error correction, and data processing onboard the satellite, the number of hops is halved, thereby reducing latency and improving signal quality. All of this leads to better performance and reliability in communication, particularly important for developing high-speed 5G applications. Additionally, more compute power means that Regenerative architectures can allow for more complex operations such as network slicing directly in space, offering greater network flexibility and scalability
Finally, the Regenerative architecture also provides the option to future proof LEO satellites, with the possibility of upgrades to 6G in the future. While 6G standards haven’t yet been finalised, it is expected that it will fit into the Regenerative framework - with migration likely to be a software upgrade to existing base stations. This means that NTNs architected using this approach will likely provide a clear upgrade path for the future, extending both their capabilities and lifespan.
Conclusion
While it’s clear that both approaches offer advantages, we believe that the benefits of the Regenerative approach significantly outweigh those of the Transparent architecture.
Although the simpler Transparent approach has the advantages of lower power requirements and being slightly more cost-effective to deploy in the short term, the Regenerative approach only incurs a relatively small cost penalty, and the higher nominal power is more than compensated for by greater performance, flexibility, and an upgrade path.
These features make it a much more attractive option for advanced services, both now and in the future.
Author details: Eric Dowek, Director of Marketing, AccelerComm
