Non-terrestrial networks (NTNs) have served remote users beyond the reach of cellular networks for years, but until now have rarely been integrated with mobile services. Growing demand for ubiquitous global wireless connectivity has grown with the rise in high performance mobile services and expectations of access wherever users or devices are, whether during transportation, in unserved areas for broadband, or for a growing array of IoT applications connecting devices in the field.
The mobile industry has been relatively slow to embrace NTNs, only starting to grapple with the challenges after first standardization of 5G in 3GPP Release 15, published in mid-2018. Since then, through Releases 16 and 17, 3GPP has been exploring how best to embrace all NTN options and aspects within a common 5G framework that takes account of the different dynamics from terrestrial cellular operation.
This work has crystallized around four areas that we identify as technical, CPE or end device, regulatory or legal, and definition. This last area is the starting point because, until there are clear definitions of distinct NTN categories, it is impossible to address the challenges logically.
It is worth considering that terrestrial networks themselves can be split into sub-categories, those where the CPE can be mobile and those where it is always fixed. This distinction is important because when the CPE is always stationary, as in the case of fixed wireless access (FWA) services, the antenna can be designed differently and enable greater overall capacity by a factor of at least three and potentially up to 10.
There also significant distinctions between categories of NTN. There are different use cases, varying abilities to reach users or devices beyond the reach of cellular, and there is provision of back-up to terrestrial networks to cater for loss of infrastructure for various reasons, such as earthquake or war. There is also interest in use of NTNs for FWA, using fixed aerial access stations to extend coverage in some rugged remote areas more cost effectively than with ground base stations. Aircraft services also constitute a distinct case, served by a combination of ground base stations having upward facing antennas, satellite, and even the planes themselves.
The general definition of NTN revolves around base stations being airborne, or non-terrestrial, but even there overlap is occurring in the case of FWA. This then leads to two broad NTN categories, satellite and High-Altitude Platforms (HAPS), usually in the stratosphere. These in turn are sub-divided, into LEO (low earth orbiting), and GEO (geostationary) for satellites, while HAPS can be balloons, autonomous gliders, or powered drones.
This leads to technical challenges, with a common element being increased distance between client and base station to varying degrees, compounded in the case of satellite and aircraft by the speed of motion during transmission. This speed results in large Doppler shifts, increasing or decreasing the effective RF frequency during transmission, requiring estimation techniques to calculate the adjustment that should be made during receipt of signals. Additional signal processing is also required to ensure reliable handover when client devices are fast moving, or between terrestrial and satellite transmission.
At this stage, neither 5G nor NTN networks are ready for full integration, which will await 5G-Advanced, in Releases 18 to 20, starting from late in 2023 at the earliest.
So far, 5G and relevant higher performing NTNs have emerged as islands with separate technology, even if use cases and business models increasingly overlap. This will also await a new generation of HAPS-based networks, beyond the initial projects such as Google’s Project Loon, which staked out the ground and identified future potential, but failed to enable competitive services. Similarly, it will await maturation of current satellite communication ventures funded by multibillionaire tech entrepreneurs, such as Elon Musk’s Starlink operated by his SpaceX, and Jeff Bezos’ Blue Origin.
On the satellite front, there is also a need to cater for two separate frequency ranges, the S-band in the 2-4 GHz range, and Ka-band at an order of magnitude higher in the 26.5-40 GHz range, as well as varying elevation angles and signal distances. Some challenges of terrestrial communication are reduced above ground, as in multipath fading, where there is less or no need to cater for objects along the signal paths, which at the distances of satellites are effectively parallel and so do not intersect.
Certainly, parameters governing signal transmission, primarily line of-sight probability, angular spread, antenna elevation and delay spread, are quite different in NTN compared to terrestrial networks.
One point to note on the standards front is that NTNs will serve not just the higher performance and demanding 5G use cases, but also lower power IoT cases that call for massive machine type communication (mMTC) capabilities to support large numbers of small bit-rate transmissions. To cater for this, 3GPP is also working on adaptations of the two cellular LPWAN protocols, NB-IoT and LTE-M, to support NTN in Release 17.
A distinct positive is that NTNs will be mesh networks, often comprising different types of node including satellites and aircraft, as well as ground base stations. This will confer potentially greater coverage, scalability and robustness, as well as improving the performance of 5G terrestrial networks in unserved and underserved areas. It could also ensure continuous service for IoT devices in the field, as well as mobile handsets on-board ships, aircrafts, and trains.
The CPE aspect of NTNs is closely related to the technical area, embracing device form factors, cost, and power levels. A starting point here is that currently quite different terminals are required for NTN communications, because about 10 times more power is needed to transmit signals to the nearest satellite. That is why satellite phones are larger and heavier, more resembling early consumer handsets. The first integrated handsets would need to be dual-mode and probably would not be as convenient to use or carry. Improvements will then come through power efficiency, further component miniaturization and above all antenna design. Direct satellite connectivity for standard handsets is a key area of industry research and a pioneer is AST SpaceMobile (see separate item).
Finally, on the regulatory or legal front, one challenge follows from the fact NTNs by definition cover multiple countries, or can do, just like direct-to-satellite TV. This will be new for many MNOs, which will have to ensure their services conform with regulations over operations and data gathering in multiple countries, although that ground has been covered to some extent by satellite platform operators and their service provider customers.