Inflight connectivity is a keystone of aviation’s digitalisation agenda, and for the most part that means satellites rather than air-to-ground networks. With thousands of new satellites being lofted into orbit, what are the differences between geostationary, low and medium earth orbit satellites — and what they can do for airline operations and passengers? Join us for the latest in our ab initio series of primer guides.
At the most basic level, geostationary earth orbit (GEO), low earth orbit (LEO) and medium earth orbit (MEO) satellites are similar in that they communicate using radio waves with antennas on top of airliner fuselages. They differ, though, in four main interlinked ways: the distance they orbit the earth, the size of the geographical area that they can therefore cover, the physical size of the satellite itself, and the capacity designed for the satellite based on the constraints inherent to the previous factors.
- GEO satellites are the largest, and orbit at 36,000km
- MEO satellites are in the middle, orbiting at 5,000–20,000km
- LEO satellites are the smallest, orbiting at 500–1,200km
In terms of size, before their solar power panels unfold, GEO satellites are upwards of the size of a bus, MEO about the size of a small car, and LEO satellites roughly the size of an armchair.
GEO satellites, as the name suggests, are positioned in high orbit at the precise point where they orbit the earth once a day, and thus they appear stationary above a specific point of the earth. MEO and LEO satellites, by contrast, are lower in orbit and circle the planet in a matter of hours. In inflight connectivity terms, this means that a provider needs fewer GEO satellites to achieve global coverage, more MEO satellites to do so, and even more LEO satellites.
To use an analogy, imagine you have a big beachball (the earth) in a dark, empty room (space), and three flashlights (satellite beams): a large, powerful police-style flashlight about 30cm/1ft long, a medium household flashlight measuring 15cm/6in, and a small keyring flashlight about 8cm/3in.
If we stand well back from the beachball (in a geostationary beachball orbit) we can shine light (here representing connectivity) strongly onto a full third of it with the big flashlight. To achieve the same brightness with our small keyring flashlight we need to move much closer to the beachball (a low beachball orbit). For our medium flashlight we end up somewhere in the middle (a medium beachball orbit).
Since GEO satellites can cover a larger geographical area, it makes sense for them to be larger and more powerful (in our analogy, to be a larger and more powerful flashlight), with more antennas and beams, and the inverse is thus true for MEO and LEO satellites (or flashlights). This size, however, means that they are individually much more expensive to build and launch, and that the impact of any one issue — such as the problems with the ViaSat-3 F1 satellite, which will operate at just ten percent capacity — is much greater.
For basic global coverage with a GEO satellite, a provider needs just three satellites, which are usually orbited over the equator since fewer people live (and fewer airplanes fly) at polar regions. MEO and LEO satellites need many more satellites to achieve global coverage, since they cover less of the earth and need to take into account that they are moving in and out of coverage areas.
In real-world examples, GEO networks from Viasat and Inmarsat (later acquired by Viasat) required just three initial satellites to achieve global coverage, with later launches adding regional capacity where required.
At the MEO level, SES’ O3b connectivity network orbits at just over 8,000km and uses 20 satellites, with 11 supplementary O3b mPOWER satellites launching between 2022 and 2024, focussing on the more populated areas of the planet.
OneWeb’s LEO constellation, meanwhile, has more than 600 satellites deployed, with Starlink’s satellites currently numbering over 5,000 and expecting to grow to 12,000 at present and potentially 42,000 later.
Most inflight connectivity satellites use either Ku-band or Ka-band antennas. These offer variable performance based on what the connectivity provider can offer, the region of satellite coverage, and what the airline has chosen as its service level in terms of speed and bandwidth. Performance is often better over populated land areas, since aviation is usually just one of the customer segments for satellite connectivity, and providers therefore concentrate the antennas and beams over land — to be used, for example, to provide residential or industrial connectivity for rural or remote locations without landline or mobile Internet.
By and large, though, in the aviation context Ka-band tends to outperform Ku-band in terms of end user experience, although this performance tends to be more relevant to passenger use than operational flight deck or cabin crew purposes. To perhaps oversimplify, it’s a big deal for a passenger if an Instagram story takes thirty seconds to load, but not such a big deal if it takes thirty seconds for a flight attendant’s iPad to send an inflight catering restock request to a provisioner.
A further factor between GEO, LEO and MEO affecting real-world performance on the aircraft is latency, often referred to as lag. At its most simple level, a basic Internet request might travel satellite-aircraft-ground station-satellite-aircraft. Even at the speed of radio waves, the sheer distances involved — hundreds of thousands of kilometres if bouncing just a couple of requests back and forth — can result in what users perceive as a delayed reaction.
In numbers, the average domestic or business broadband connection might have latency (measured as ‘ping’ using speed tests) of 10 milliseconds. Your author undertook a recent GEO satellite connectivity speed test on a brand new Airbus A350 that measured 1,796 milliseconds. The real-world performance felt correspondingly sluggish when browsing or using social media, but when streaming videos the effect is less noticeable. With their lower orbit altitudes, MEO — and especially LEO — satellites have significantly lower latency, giving a much stronger real-world performance.
Looking to the future, hybrid networks are clearly in focus for most providers, taking advantage of the benefits of all three kinds of satellite: splitting requests at the aircraft between latency-sensitive requests going to LEOs and non-latency-sensitive requests going to GEOs, for example, or adding multiple layers of service at congested areas. A complex balance of constellation, band and onboard technology will be required, but the future Internet performance onboard will be substantially better than it is today.
To read more on this topic, see the recent article Yocova on Air: Better inflight connectivity and satellite technology – Hughes Network Systems.
Author: John Walton
Published 16 November 2023
;
