The impact of Covid on connectivity has been wide reaching, shining a spotlight on digital inequality not only between countries, but within countries. It has also exposed the need for continued (if not increased) efforts to improve access and adoption as a large majority of the global population has in some way shifted towards remote work, education, entertainment and healthcare.
Connectivity remains a problem as the ITU Broadband Commission’s most recent State of Broadband Development Report (2020) found that 46.4% of the world’s population currently remains unconnected.
Enter … satellite broadband.
Over the past decade, the satellite industry has gone through significant technological evolution. This has resulted in lower barriers to entry and a new cast of characters such as Space X, OneWeb and Amazon (Blue Origin) that have emerged to challenge the “good ole’ boys club” of traditional satellite operators.
The game changer for these new entrants has been the fully reusable rocket, which has substantially reduced the cost of space access and is the key to the deployment of large-constellations of small satellites in low Earth orbit (LEO) to deliver telecommunications and internet services.
As a refresher: Satellites are typically deployed into three key orbits: low Earth, medium Earth and geosynchronous.
The key differences between the different orbits is altitude, latency and coverage. The further away, the longer it takes a satellite to revolve around the earth. For instance, it takes 24 hours for a GEO satellite to revolve, while MEO range from 2 to 24 hours. Due to its much lower altitudes, LEO satellites take about 90 minutes. From a latency perspective, GEO is around 500ms, while LEO is under 50ms. Finally, due to the fact that GEO is typically placed in a fixed position, its coverage is reduced but more targeted, while LEO satellite constellations are able to provide continuous and uninterrupted coverage over a significantly wider geographic area.
SpaceX Starlink shines bright
Proving that things are only impossible until they’re not, SpaceX has had a banner year. With its Starlink satellite broadband services, SpaceX has made major strides against its competition. In the past year, SpaceX has launched 25 Starlink missions that have placed 1,565 spacecraft into orbit to beam broadband signals around the world – nearly nine times as many satellites as any other company’s constellation. The company has FCC authorization to deploy 12,000 Starlink satellites operating at Ku-band, Ka-band, and V-band frequencies, and at a range of altitudes and inclinations in low Earth orbit to support its beta customers. Its “Better Than Nothing” beta service is currently supporting 10,000 subscribers paying USD 99 per month for a service that provides speeds of 50Mbit/s to 150Mbit/s and latency from 20ms to 40ms. More importantly, Starlink has stated that it currently has over 500,000 pre-orders. As more spacecraft are launched and coverage is expanded, speeds are expected to reach up to 300Mbit/s.
To put that in perspective, current total global satellite broadband subscribers are around 2.6 million. And to further put this in perspective, most satellite broadband services are typically in the 25Mbit/s downstream range.
To date, Starlink is far ahead of its competitors:
- OneWeb was acquired by the UK government and the Bharti Group in October 2020. It has launched 182 of its planned 650 satellites and is on schedule to cover 50 degrees latitude and above by June, with service ready to start by the end of the year.
- Telsat plans a fleet of 268 LEO satellites, with the first satellites to be launched at the beginning of 2023, with customer beta testing beginning shortly thereafter and commercial services commencing in the second half of 2023.
- Amazon’s Project Kuiper, with a 3,200-satellite constellation planned, has not launched a single satellite.
LEO vs. GEO
Many of the incumbent GEO operators are planning to complement their satellite portfolio with LEO satellites. In their view, GEO will remain the most efficient way to deliver content to large geographic areas particularly with regards to broadcast video distribution, while LEO satellites can offer the better latency and the lower-cost bandwidth from GEO satellites making it highly attractive for applications supporting maritime, aviation and government.
However, with the introduction of very-high throughput satellite (VHTS) systems, these traditional operators believe they can effectively compete in both price and throughput with these new entrants. Eutelsat, one of the largest satellite broadband operators globally, is sticking with GEO. It currently offers its Konnect satellite broadband internet services in the UK and 15 other countries in Europe and North Africa, offering download speeds up to 75Mbit/s and upload of 3Mbit/s. It is expected to launch the KONNECT VHTS satellite this year with service available in 2022 with expanded access across Europe.
ViaSat currently operates five GEO satellites and will be putting three "ultra-high capacity GEO satellites" into orbit at the start of 2022, which will offer global coverage by 2023. However, it plans to complement its GEO satellites by putting 288 satellites into low-Earth orbit by 2026.
Hughes Net is sticking with its GEO, however it is complementing this strategy through its investment in OneWeb (via its parent company EchoStar).
With over a half-million people on the waitlist for Starlink, it feels like Elon Musk has transferred the lure of his Tesla cars to the broadband internet space.Time will be the true test of success
The current success or perhaps “hype” of Starlink reminds me a lot of the early days of Google Fiber. Everyone wanted to be or have part of it. With over a half-million people on the waitlist for Starlink, it feels like Elon Musk has transferred the lure of his Tesla cars to the broadband internet space. It may even be a package deal – buy a Tesla and get Starlink service. Nonetheless, time will be the true test of success and whether or not it can live up to expectations.
According to the Ookla speed tests for 1Q21, median downstream speeds ranged 40.36Mbit/s 93.09 Mbit/s in the US. In Canada, median downstream speeds ranged from 53.61Mbit/s to 80.57Mbit/s. Not exactly “out of this world” speeds – but then again … it is called the “Better Than Nothing” beta for a reason.
From a latency perspective, median latency values in the US were observed from 31ms up to 88ms. This compared to fixed latency values of 8ms to 47ms. In Canada, median latency faired a bit better with rates of 34ms to 61ms. In both cases, latency values were outside the expected range.
Concern remains that the sheer number of satellites in LEO will cause orbital congestion with the potential for satellite collisions and the creation of space pollution.
On the business side, these projects remain capital intensive and will likely be challenged to meet projected cost targets – especially when the target markets are especially price sensitive. This is particularly true as it relates to the costs of the terminal and other necessary equipment.
Finally, service delays could also impact the business models and we have already seen significantly delays in satellite launches from many of the operators.
Although the current beta results from Starlink are nothing astronomical in terms of speed and latency, it does show that LEO-based satellite broadband could dramatically improve broadband access in difficult-to-serve areas without cost of laying physical infrastructure. It also proves that these projects could help to close the digital divide by offering fixed-broadband-like speeds; enable mobile network operators to expand wireless services; as well as provide high-throughput mobile broadband connectivity for aircraft, maritime and land vehicles.