Advancements in satellite launch costs, latency and capacity and a huge spike in capital spending by satellite and investors, is changing the way MNOs think about the use-cases for non-terrestrial networks – and towercos must adapt.
What are non-terrestrial networks?
The non-terrestrial space can be broken up into four segments, three at the orbital level and one at the stratospheric; each has various use-cases and advantages.
The furthest afield are Geostationary Equatorial Orbit (GEO) satellites, which sit at an altitude of 35,786km above sea level. GEO satellites are a type of geosynchronous orbit (GSO) satellite that always remain stationary over a single point in the sky and are mostly used in telecommunications and for weather monitoring.
Moving closer at under 35,786 and above 2,000km above sea level are the Medium Earth Orbit (MEO) satellites. These, unlike GEO satellites, orbit the earth and are most used for navigational purposes due to their ability to cover large areas and move freely around the earth.
The closest are the Low Earth Orbit (LEO) satellites at an altitude of between 2,000 and 160km. These cover much smaller areas of the earth but at a much closer range.
GEO satellites are more expensive to deploy and harder to replace due to their extreme altitude. This distance also reduces latency, limiting the amount of capacity they can support. LEO, on the other hand, supports much greater connectivity speeds due to its closer proximity to ground-based handsets and base stations. MEO offers an in-between, serving a middle ground between both GEO and LEO, but also not maximising on either advantage.
Finally, entering the atmosphere we then have high-altitude platform stations (HAPS), also known as stratospheric satellites. Rather than remaining in orbit, these sit within the stratosphere. HAPS are mostly in the pilot-stage and see use for rapid deployment in response to any downtime of the terrestrial network, for example in cases of natural disaster.
As HAPS sit closer to the ground than even LEO, they offer more use-cases for voice and data services, making them a great temporary solution for full cellular services, but there are limitations to how long HAPS can remain in the air and are more expensive than any orbital coverage network. A HAPS is most similar to a cell tower given it allows cellular coverage of a particular rather than always being on the move.
Supporting backhaul connectivity
The bridge between NTN and terrestrial networks has been strongly established in backhaul service provided by satellite operators for MNOs and towercos. In rural areas, the use of fibre or microwave backhaul for cellular and broadband services is often impractical.
The combination of high capex for fixed backhaul network deployment and low opportunities for revenue generation usually make any return on investment (ROI) non-existent. This can be one of the biggest costs associated with running rural terrestrial networks, particularly in difficult terrain such as mountains, valleys or heavily forested areas.
Satellite backhaul allows for the deployment of 4G, 5G and IoT coverage via a series of integrated terrestrial and satellite networks. Satellites can be used to link backhaul between data centres and hubs located near the core of their customers’ networks, to ground stations on towers in rural areas.
Intelsat is working with rural network-as-a-service provider AMN to deliver backhaul services for 500 of their sites in Madagascar for mobile broadband access. The combination of satellite and solar powered base-stations has helped create a tower solution that can bring connectivity to the most remote regions of the earth.
This has two distinct benefits for towercos. One is that it makes the business case for the deployment of rural sites more commercially attractive to MNOs, generating demand for more tower builds.
Satellite backhaul providers can also improve speed of site deployment, as the microwave or fibre network is often the last items to be delivered on a tower build, while satellite backhaul can be near instant; a matter of days or even hours. IoT monitoring systems via satellite backhaul mean towercos can also receive instant visibility over the site.
Keeping networks resilient
Satellite connectivity can also act as an extra layer of network resiliency if there was to be any disruption to a terrestrial network. In Africa, loadshedding, theft, extreme weather conditions and conflict can all cause terrestrial backhaul networks to go down for periods of time that could last from a few hours to days or even weeks.
This was seen prominently in South Africa, as the decline of the state-owned energy grid Eskom led to large spikes in loadshedding and network downtime for operators over the last year. The South African Reserve Bank urged local banks and operators to improve their resiliency to ensure critical financial systems remain active, which was part of a 2022 review.
Satellite operator Avanti has been increasing its focus on the market by providing satellite resilience to the financial sector, providing backhaul between data centres across the country, and even to some in Europe should local infrastructure be restricted due to power outages or natural disasters. Some financial centres in Africa route their data processing through South Africa, so any downtime can cause issues for the entire continent.
While resiliency is a great challenge in Africa overall, natural disasters are a threat in any market exposed to extreme weather – a growing problem due to climate change. In 2017, Hurricane Harvey tore through the Gulf Coast in the US knocking out 70% of cell towers in addition to 200,000 homes. Responders needed alternative connectivity measures to stay active, and Viasat was able to provide temporary satellite communication services.
Towerco opportunities in ground-based stations
Satellite backhaul also opens opportunities for a new adjacent service. Satellite ground stations are a core part of satellite networks, acting as a terminal for signals to transmit from terrestrial to non-terrestrial. These ground stations vary massively; from huge dish antennas to small V-SAT (very small aperture terminal) equipment commonly used in the telecommunication industry located on the tower itself.
This presents an opportunity for towercos to own and operate the passive infrastructure of ground-based stations, allowing satellite operators to focus their capex on their non-terrestrial network. A huge volume of these sites are needed to support dense LEO constellations, and someone needs to host all of this equipment.
Towercos are suited to operating these sites, as they already operate large, distributed networks, have experience providing resilient passive infrastructure and energy management. At TowerXchange Meetup Europe, an Intelsat executive explained that one of the biggest challenges for satellite operators is getting the right ground locations for their non-terrestrial networks.
The emerging threat of direct-to-device
Satellite operators and the towerco industry have a history of collaboration and cooperation to deliver cost-effective and resilient connectivity for operators. However, the emergence of direct-to-device services is already causing disruption, and towercos are watching cautiously.
In January 2024, SpaceX launched its Falcon 9 rocket, hosting the first set of satellites that can beam phone signals from space directly to smartphones, offering mobile network access direct to devices in the US. This will start with text messaging but is expected to follow with voice and data in the coming years.
There is a group of companies including AST SpaceMobile, Lynk and Starlink, that have ambitious plans to use terrestrial spectrum from space. In 2023, the satellite industry saw a switch of focus away from IoT to direct-to-device on the assumption that today’s emergency-text services will expand into full broadband and cellular connectivity.
These networks communicate directly between 3GPP mobile devices and an MNOs’ core networks, routing spectrum through an LEO satellite constellation and negating the need for any terrestrial infrastructure – including ground-based towers.
Currently, direct-to-device is of limited use, because it can only support messaging and only while a satellite is passing overhead. It is also considerably more expensive than terrestrial networks, relegating its use to emergency coverage and for network resiliency. Direct-to-device can support MNO regulatory coverage obligations by allowing messages to be sent and received from the most remote regions, as well as offering maritime coverage that is virtually impossible to do with any terrestrial network.
LEO constellation cost-per-units are approaching terrestrial deployment costs. SpaceX, Blue Origin and United Launch Alliance are all making launches cheaper and more frequent; a US$5-7mn price tag today is 40 times lower than it was in the 1980s and is expected to fall 95% by 2040 according to Citi Group. This massive cost reduction will boost investment and help deploy increasingly dense constellations of LEO satellites that can deliver low latency, high-speed global connectivity.
While cheaper backhauling will in turn make towers cheaper to deploy, it also makes a case for direct-to-device more compelling as well.
Satellite operators are now thinking about dedicated spectrum to deliver 5G directly, acting as a service for MNOs to reach their customers, anywhere in the world. The integration of satellite services with 5G standards is showing promise, with MNOs partnering with satellite internet players to allow smartphones and IoT devices to connect with terrestrial and satellite systems alike. In 2022, Apple introduced an emergency text feature on new iPhones using Globalstar satellites and is now financing a new generation of Globalstar satellites to launch in 2025.
The best of both worlds
While direct-to-device could be a future threat to the tower industry, for the short-to-medium term it seems the best solution is one where MNOs have a heterogenous network that combines all worlds; towers will deliver localised high-capacity coverage with peak data rates, at lower latency and cost in dense urban areas where a satellite would be incapable of handing that level of densification.
NTNs can then extend network coverage into the hardest-to-reach areas, helping MNOs meet their regulatory-mandated coverage requirements. MNOs can also benefit from having multiple levels of resiliency, as NTN networks can provide temporary coverage support in the event of any terrestrial network failure. There is no single solution to closing the connectivity gap, and demand for coverage and capacity continue to outpace competition on the supply-side.
In the long-run, a dual-pronged approach to urban and rural coverage appears the most likely path as towercos focus on high-capacity urban areas with NTNs taking over more responsibility for rural and remote coverage.
This does leave a final question as to the role of companies such as AMN, Vanu and NuRAN, who offer rural terrestrial network-as-a-service coverage.
