Satellites have been an integral part of global communications networks since the launch of the first geostationary telecoms-capable satellite – Syncom 3 – in 1964. Designed to be able to handle a single two-way telephone call or 16 one-way teletype channels, it started a revolution in global telecoms.

This was the same year as Intelsat was founded, a multi-governmental organisation which took the lead in launching and managing satellites for TV broadcast and global communications. Into the 1980s there was fierce rivalry between long-distance submarine cables and satellite for global voice provision, with submarine cables pulling ahead following the introduction of optical fibre transmission and now well ahead due to the massive capacities available and needed to support backbone internet data requirements.

GEO was king

In the early days communications satellites stayed where they were put, they tracked above the same spot on the earth meaning an antenna need not move to follow them. However, this put them around 36,000 km above the earth and more importantly to telecoms services gave them a latency of close to 600ms for a round trip. This delay was clearly audible though just about tolerable for PSTN voice calls, but would be way above the maximum recommended nowadays for services such as Skype which suggest no more than 100ms round trip from end device to skype node – and this to include all of the processing in terminals, LANs and local networks.

MEO and LEO – the new frontier

To solve the latency problem satellites have to be lower, but then they move relative to the earth so need sophisticated tracking capabilities to use them at high bandwidth. Advances in ground station and terminal technology has enabled the Mid Earth Orbit (MEO) and Low Earth Orbit (LEO) satellite “constellations” to emerge.

There is no hard and fast boundary between LEO and MEO however an altitude of around 2000km is assumed with MEO being anything between this and the GEO at 36,000km and LEO being anything below.

Perhaps the most famous MEO satellite was Telstar 1 – the first ever experimental communications satellite launched in 1962.  Many military and Global Positioning Systems (GPS) satellite networks use MEO orbits, as does the communications service O3b which describes itself as “the world’s only commercially successful non-geostationary satellite system”.  O3b started in 2010 with the aim to bring satellite internet services and mobile base station backhaul services to emerging markets – to address the “Other 3 billion” people in the world without access to terrestrial broadband internet.  With a fleet of around 20 satellites at around 8000km, O3b’s parent company SES are about to launch the second generation called O3b mPower which will be able to deliver anywhere from hundreds of megabits to 10Gb via use of 30,000 spot beams.

The dawn of the LEOs

However, the latest development in satellite technology, and perhaps that of most interest to service providers and corporations considering global WANs, are the LEO constellations. One of the first in this space (so to speak) was Iridium, who have a network of 66 satellites at a height of around 1400km which gives a round trip delay of around 20ms – however as the satellites move the latency changes depending on whether one is directly overhead or at an angle. As an added complication Iridium satellites communicate one to another rather than always straight-up and straight down. Iridium already provides a host of specialised communications services to quite specialised applications such as maritime, aeronautical, government and land mobile services.

LEOs for the masses are already with us though – through the likes of Starlink and OneWeb.

Starlink and OneWeb

Starlink is the brainchild of technology entrepreneur Elon Musk, an offshoot of his SpaceX rocket launch capability.  In January 2020 Starlink became the largest satellite constellation in the world and is currently offering its beta trial service – broadband internet at speeds between 50Mb/s and 150Mb/s and latency of 20ms to 40ms. These statistics are expected to be improved upon as the network of satellites and earth stations is expanded. The target market is direct sale broadband internet services – a direct competitor for terrestrial broadband for those locations in developed countries (such as the USA) who can’t get fast reliable copper/fibre-based internet services.

OneWeb is perhaps most famous for going bust, but the UK-based company has secured funding from the UK Government and Bharti Global and is now building its fleet and by May this year had launched around 220 of its planned 650 satellite constellation. Operating at an altitude of around 1200km, OneWeb promises “ten times the bandwidth and one tenth of the latency” of GEO satellite solutions.  Its target customers are primarily businesses, governments including defence, phone network operators and clusters of communities. The OneWeb introductory video is worth a look as it gives a good overview of how the service will be provided and accessed – you can find it here.

Impact on Global WAN provision

With the ubiquitous connectivity and low latency, all global WANs will migrate to the LEO offerings, won’t they? Well, whilst an interesting technical offer, the main issue here will be price which is anticipated to be significantly higher than terrestrial connectivity solutions. Service providers and corporates are likely to find themselves able to access these services indirectly for those sites that just can’t be reached any other way – for example OneWeb have just announced that they have signed an MOU with BT “to explore the provision of improved digital communication services to some of the hardest to reach parts of the UK” – in anticipation of OneWeb capability becoming an option for broadband provision within BT’s internet portfolio.

Perhaps deviating from its “broadband to the masses” ethos, Starlink have announced deals with the likes of Microsoft azure – Microsoft partners with SpaceX to connect Azure cloud to Musk’s Starlink satellite internet enabling the kind of direct access to hyperscale hosting capability that will be attractive for business as they move to the cloud.

An attractive offer (depending on price) to a service provider in the global WAN business would be flexible use of a set capacity on the system – for example leasing a permanent 100Mb of capacity from anywhere to anywhere – meaning that this could be used for very fast turn-up of new locations, or for disaster recovery, and then re-used once a more cost-effective terrestrial solution has been put in place.

And into the future

Taking a leaf from the Iridium book, future generations of LEOs can expect to add satellite to satellite communication and on-board processing and routing.  These “routers in the sky” would connect most likely by laser and provide very low latency point to point capacity which would beat the latency of terrestrial alternatives for inter-continental traffic – but of course for a price. This would in effect completely change the image of initial geostationary satellites from high latency, inflexible systems to become the offering for ultra-low latency into for example the wholesale finance vertical, and an essential stop-gap offering for rapid turn-up in the portfolio of all global WAN providers.

Here at Brodynt we will be keeping a close eye on LEO developments especially as and when some of our existing ISP suppliers integrate them into their internet access networks for the hard-to-reach locations. This is yet another technological development which will enhance the attractiveness of the internet to provide global WANs, even to the remotest of locations.

You are always welcome to drop us an e-mail to hello@brodynt.com and one of our Account Managers will be happy to assist you! Alternatively, engage in a discussion by leaving a comment below and don’t forget to share the article with your friends and colleagues!