Separating Hype from Viability: The Role of LEO Satellites in Mobile Networks

Telesat and Telefónica International Wholesale Services (TIWS) have recently completed a live in-orbit trial of a broad selection of applications on Telesat's Low Earth Orbit (LEO) Phase-1 satellite.

Applications tested over Telesat LEO resulted in observed round-trip latency of 30-60 microseconds without any packet loss. Test scenarios included:

• High definition video streaming
• Video conferencing
• Remote desktop connection to manage a remote device
• Virtual Private Network (VPN) connection
• Bi-directional File Transfer Protocol (FTP) encrypted file transfers of 2 GB
• Internet Protocol Security (IPSec) tunnel encryption with no reduction in the performance of the link

Latency is the main benefit that LEO satellites bring to mobile networks. Orbiting at altitudes between 400 km and 1,500 km, LEO satellites have a latency period of 20 to 60 microseconds, a much lower latency rate compared to the 140 microseconds and 280 microseconds latency rates of Medium Earth Orbit (MEO) and Geostationary Earth Orbit (GEO) platforms respectively. The trials of Telestat’s LEO satellites have exhibited a substantial improvement in latency performance over GEO links without the conventional use of compression or TCP acceleration techniques that are typically required in high latency GEO environments. TCP Acceleration involves using a set of enhancement technologies that speed up underlying communication protocols (TCP and HTTP) so that the user will have a better Quality of Experience (QoE).

Another distinct feature of LEO satellites constellations is that they are deployed on a much larger scale relative to MEO and GEO to provide ubiquitous global coverage. Lower orbit altitudes require more satellite units to be deployed; notable LEO satellite provider SpaceX was approved for 12,000 LEO payloads and has applied for the permission from the Federal Communications Commission (FCC) to launch 30,000 more satellites in the medium to long term. Conversely, MEO satellites require 8 to 20 satellites while GEO constellations require only 3 satellites for an expansive global footprint.

There are firm grounds for industry optimism toward LEO satellites communications. LEO constellations can provide omnipresent connectivity to rural communities that have capacity rates (touted to be at gigabit speeds) capable of rivaling the speeds of terrestrial networks. LEO satellites, used in standalone or for redundancy purposes, would also enhance reliability rates for latency-sensitive, mission-critical use cases (especially in completely or remotely rural areas) that 5G can bring forward.

However, this bolstered optimism for LEO satellite constellations must be moderated. Despite promising trials, they are all still currently in initial development and none have been commercially deployed to support the backhaul needs of a Communications Service Provider (CSP) just yet. The timeline for LEO satellites to be an established fixture in an operator’s arsenal of backhaul solutions should account for how LEO satellites can overcome the commercial challenges of mass manufacturing and deployments.

By virtue of their large-scale, high-volume implementations, LEO companies must improve their business cases and overcome the cost challenges in manufacturing, ground equipment, User Equipment (UE), and launches.

The comparative manufacturing cost advantages that LEO satellites have over GEO—GEO satellites cost around US$61,000 per kg while LEO satellites can be as low as US$1,100 per kg—are negated when we account for the overall manufacturing costs that would be incurred by a LEO company in providing global coverage. High-volume satellites required for global coverage would create a large cost disparity between LEO deployments (requiring thousands of satellites) over GEO deployments (requiring less than 10). It will take time for production costs to further decline to support the viability of manufacturing the thousands of satellites that LEO constellations require.

Launch cost comparisons between LEO and GEO/MEO would also have similar outcomes. A GEO satellite launch would cost about US$30,000 per kg while LEO launch costs would only be about US$5,000 per kg—a cost disparity that would be upended when overall launch costs for the total number of satellites within each respective orbital type are accounted for.

ABI Research forecasts that LEO satellites will establish themselves as a prominent fixture in a CSP’s portfolio of backhaul solutions only in 2023, accounting for a conservative 2% of overall satellite backhaul links (with the rest of the backhaul links being supported by GEO and MEO satellites).