TelcoFuturism - Part 5 Satellites
This is the fifth blog post in a series on “TelcoFuturism” – which focuses on the intersection points between the telecom industry, areas of advanced technology, such as drones, AI, AR/VR and robotics, plus societal changes such as climate change & shifting demographics. Disruptive Analysis looks at these adjacent technologies through the lens of what’s really happening in networking, opportunities challenges that could emerge and the tough practicalities and complexities telcos face, rather than accepting the general hype and rose-tinted views common among some forecasters.
Satellite: a new era in orbit
The satellite sector is not new. For decades it has enabled broadcast services, international and maritime communications, scientific sensing and much more. Currently, the global industry is worth about $11bn/year, with around 2000 operational satellites, mostly in “geostationary” orbits 36,000km above the Earth, which keeps them static over specific locations.
The telecoms industry has long had tight links with a satellite sector, including sharing the ITU (International Telecoms Union), a UN agency, as the main regulatory body which covers aspects such as spectrum allocations and orbital permissions. The original 1960s/70s INTELSAT system had national ITU signatories as members, and provided mostly international voice communications services for the telecom industry initially, before expanding to broadcast and data sectors. In subsequent decades, the market has been privatized and expanded significantly.
Today, most telcos have minimal involvement in owning and operating orbital resources directly, but many run their own earth stations (control and uplink centres) for various purposes, and use satellite connectivity for capabilities such as
- Backhaul for mobile base stations in remote locations, such as islands, mountains and other sites far from fibre or microwave reach.
- Distribution services for broadcasters
- Satellite TV as part of bundles – for instance AT&T’s purchase of DirectTV a few years ago.
- Integration into custom solutions for industry verticals, for example in shipping or public safety.
But the industry now has lots of further change coming in the next few years, not least with various Internet giant companies (or at least their founders) becoming space entrepreneurs. This has both positive and negative implications for terrestrial telecoms, whether that relates to wireline, wireless or transport segments.
Satellite industry evolution in the 2020s
The fundamental industry economics of satellite are changing fast. Historically, it cost around $450m to launch one, giving the owner an asset with a 15-year lifespan to earn a return. This was made up of, approximately:
- $250m for the satellite itself
- $100m for insurance
- $100m for the launch and rocket
This economic structure tends to limit satellite connectivity to high-margin, low-volume applications – although that is also partly governed by the cost of service providers’ “earth stations” and the need for end-users to have large antennas/dishes. The frequencies used also limit signal reach through walls or roofs, meaning that direct connection to devices isn’t possible without line-of-sight.
But satellites are evolving fast, with three particular trends of importance here:
- SmallSats, which as the name suggests are lower in weight than traditional versions - typically 100-500kg, compared to 1-5 tonnes in the past. This enables multiple satellites to be launched simultaneously, resulting in a lower cost for each. (Even smaller “CubeSats” can be much lighter still, but don’t generally have the power to be useful for communications applications).
- High-Throughput Satellites (HTS). Previously, most communications satellites have provided total network capacity of around 6Gbps, usually shared over wide areas of the planet from geostationary orbits. But this is now increasing significantly, with improved antenna technology enabling “spot” beams, and some units being designed for 40-500Gbps, and eventually even up to 1TBps. This is greatly reducing the cost per transmitted bit, and coupled with lower latencies from LEOs (Low Earth Orbit) has the potential to redefine the traditional broadband satellite access model.
- LEO Constellations. Perhaps the most exciting development is the imminent creation of some huge multi-satellite constellations at low orbits, some with 100s or even 1000s of “birds”, travelling in complex meshes across the entire surface of the planet. These have the potential to enable mass-market broadband connectivity, as well as mobile backhaul for remote areas, and new classes of IoT and vehicular connectivity. The lower orbits also mean less power is needed on the end-devices, making battery-powered handhelds more realistic.
In addition to the “birds” themselves, SpaceX has now created reusable (and fairly reliable) rockets, reducing launch costs and potentially insurance. Other rivals are also pushing reusable or cheaper launch vehicles, such as RocketLab, Virgin Orbit and Blue Origin. Greater use of software also means that satellites’ functions and beam dynamics can be adapted in-life, which can also increase potential returns for their owners.
Starlink already has around 200 satellites in orbit, and may ultimately have as much as 23Tbps of aggregate capacity. Others such as OneWeb and Project Kuiper are also promising – the former having already conducted launches.
The role of advanced technology companies
One notable feature of the current satellite industry is the presence of major technology/Internet firms. Both the companies themselves, and in some cases their founders or executives personally, see the sector as enabling a range of new capabilities from “connecting the unconnected” through to enhanced cloud services provisioning.
Among the notable participants are:
- StarLink, which is being developed by Elon Musk’s SpaceX business. It is aiming to launch as many as 12000 satellites by mid-decade.
- OneWeb, which has investment from SoftBank and Qualcomm, and which has several ex-Google staffers closely involved with it. It has launched a small number of initial satellites, but is intending to ramp up to about 650 in total in coming years.
- Blue Origin (rockets) and Project Kuiper (LEO constellation) which are owned by Amazon’s Jeff Bezos. Amazon has also launched AWS Ground Station, which allows companies to “Easily control satellites and ingest data with fully managed Ground Station as a Service” – and points to a tight future integration with its other cloud / XaaS interests.
- Google has also been working with NASA on various projects, as well as having interests in (non-space) “High Altitude Platforms” such as its Loon balloon project.
- Facebook also recently received FCC permission for an initial satellite launch, and is understood to have plans for an extensive constellation of its own, although details are scarce.
- Apple is known to be interested in the satellite domain, but it is unclear if this is a major initiative such as a LEO constellation of its own, or a possible interest in integrating network capabilities into future devices.
It is worth noting a comparison with subsea fibre here – something that long used to be a domain of telco-led consortia, but is now becoming dominated by major Internet and cloud players’ investments.
Satellite & the 5G era: reality or politics?
If you have attended a 5G conference in the last few years – especially one covering spectrum or regulatory issues – you may have been surprised by the numbers of presentations about satellite communications. Most have had a similar message: “Satellite networks can integrate with, and support, 5G”.
There are actually two separate objectives at work here:
- Firstly, satellite communications actually can enable a wide range of interesting hybrid scenarios with terrestrial mobile communications. Some of these are covered in depth below.
- Secondly, the satellite industry is very keen to hold on to as much of its current global spectrum allocation as possible, in the face of determined efforts by the cellular and WiFi industries to give up large slices of it. Positioning itself as an important, innovative adjunct to 5G – especially in front of telcos, policymakers and investors – is a wise strategic move, from a lobbying standpoint.
The satellite/mobile integration has some important aspects. While the potential roles and benefits of satellites have been studied in earlier 3GPP standards releases – and satellite backhaul is used in remote locations such as connecting islands and ships – this is now receiving greater focus, especially as 5G aims to support various new capabilities such as industrial and mission-critical connectivity.
3GPP’s forthcoming Rel17 includes work on both 5G NR & enhancements of 4G NB-IoT to support satellite for key cellular use cases. The central idea is that space-based connectivity “extends the reach” of 5G. For instance, satellite backhaul can support certain verticals in areas with no fibre or reliable coverage:
- Healthcare, such as home treatment in rural areas, connected ambulances etc.
- Temporary sites such as music festival, construction sites etc.
- Reinforce service-continuity for devices and vehicles (e.g. passenger cars, aircraft, ships, high speed trains, buses)
- Enable true ubiquitous coverage especially for critical communications, future railway/maritime/aeronautical communications and truly-nationwide M2M-type uses such as utilities
The growing maturity of small cells, together with ubiquitous backhaul, has potential to change the game for the last few % of users who cannot today reliably access 4G services.
There are also interesting hybrid solutions for connecting aircraft, which can use a mix of terrestrial 5G or other ground-to-air connections for broadband, on-board WiFi, and telemetry, with satellite either for other applications, or for coverage over oceans.
Another angle for satellite/mobile integration is around broadcast of large volumes of data, for instance to many localised edge-caches or CDN nodes. A satellite overlay can be used to pre-position content for local storage, reduce duplicated data transmission needs and thus the burden on the network, and associated energy consumption. This could be imagery for gaming, updates for OS’s and other software, and similar “monolithic” datasets. While modern optical-transport networks can easily accommodate the same traffic, they may not have the “capillarity” to reach out right to the edge of the network.
There are also angles around security and resilience, for which satellite communications can dovetail well with 5G. At one level, satellite offers a backup/alternative path for both mobile and terrestrial networks, either in terms of access or for backhaul. 5G URLLC services, especially if used for critical needs such as utility grids or public safety, need to consider eventualities such as fibre cuts. Satellite could enable greater resilience. More interestingly for the mid-to-far future, quantum communications can use satellite networks for secure cryptographic key distribution.
The spectrum problem
This article is not the place for a full discussion of points of contention on spectrum that can be used for both satellite and terrestrial mobile networks. But as 5G expands in importance, there is a general focus by the industry on both midband (2-6GHz) and mmWave (>24GHz) bands.
There have already been clashes about the C-Band (generally the 3-5GHz range) in various parts of the world. The US FCC has offered to pay the satellite industry to reclaim some of its existing allocations, so they can be reused for terrestrial 5G. In the mmWave range, the concerns are more about interference with ground (or “earth”) stations, and so we may well see “exclusion zones” to protect satellite operations.
Both at the recent ITU WRC-19 conference and the initial preparations for WRC-23 in four years’ time, there have been considerable areas of disagreement, as both sectors have protected existing turf and positioned to enable new use-cases. We can expect some measures of collaboration, however – as both 3GPP mobile and satellite communities are happy to take spectrum from military, broadcast and other sectors.
Implications for wireline, optical and transport telecoms
Disruptive Analysis believes that the overlaps or competitive dynamics between satellite and optical/transport worlds are somewhat limited. Many of the use-cases are for applications or locations which cannot have fibre connections (e.g. moving vehicles), or which are so remote that it is unlikely (e.g. mountain communities).
However, domains where there are intersections between the two include:
- Links between cloud/datacentres and satellite earth stations, which will become more numerous and with higher capacity.
- Satellite delivery to edge-computing sites (e.g. for broadcast/content download and storage)
- Satellite backup as an option within SD-WAN and enterprise branch connectivity
- Local fibre/fixed broadband distribution within remote communities supplied with high-throughput satellite connections
It is important to debunk one common myth – satellite connections, even with the huge, new LEO constellations coming on stream, won’t pose a meaningful risk for normal fixed or cable broadband in urban/suburban areas of developed markets. Even with the theoretical promises of 20+ Tbps capacity (which will in any case be distributed globally) this is a tiny fraction (under <1%) of total IP access traffic. Some individual subsea cables may transmit 1Pbps (=1000 Tbps) by mid-decade, to give some context and relative scale.
To learn more about ECI’s 5G Solutions, click here.