Will geostationary satellites one day become obsolete given the proliferation of satellites in low-Earth orbit?
By Jon Kelvey|February/March 2025
In the span of six years, Elon Musk’s Starlink internet constellation has shattered the near-total dominance of large satellites in geosynchronous equatorial orbit, or GEO, as the means for delivering internet services from space. SpaceX’s Starlink subsidiary now has roughly 7,000 satellites in low-Earth orbit, and last year, it reportedly surpassed 4 million subscribers around the globe. Starlink’s success has sparked interest in LEO beyond internet services. The Pentagon is getting into the action with the Space Development Agency’s Proliferated Warfighter Space Architecture, a planned constellation of hundreds of communications and missile tracking satellites in LEO. With all the interest and investment flowing toward LEO satellites, I asked five experts if the days of GEO satellites are numbered.
Absolutely not. GEO satellites will remain an integral component of satellite networks. There are different advantages to different orbits.
Geostationary satellites are by far the most economically efficient for national or regional usage. Geostationary satellites are licensed by about 60 different nations around the world. Most of those countries use their satellites for national or regional communications applications. Often those same countries look at ownership and control of their space systems as fundamental for national sovereignty and/or security.
LEO satellite networks are inherently global and offer lower latency. But while orbital, spatial and spectrum sharing regulations are mature and successful at geostationary orbits — such regulations are virtually nonexistent on a global basis for nongeostationary orbit (NGSO) satellites in what is essentially a zero-sum game sharing environment. Such regulations are necessary not only for GEO-LEO coexistence but also for coexistence among the LEO networks of multiple nations.
U.S. domination of LEO is likely not sustainable, with U.S. licensees only accounting for about 20% of global geostationary satellite licenses and over 65% (and climbing rapidly) of nongeostationary orbit satellites. There are plans and/or pending applications for tens of thousands of additional LEO satellites, such as the planned 13,000-satellite Chinese Guowang constellation. There are substantial concerns regarding space sustainability and coordination of orbital, spatial and spectrum resources in the absence of sharing regulations.
The United Nations’ International Telecommunication Union, ITU, has emphasized in statements that the orbits and radio frequency spectrum necessary for satellites to operate are finite natural resources that must be shared and protected. Given the critical nature of sovereign space assets in the current geopolitical environment, and the economic superiority of modern high-throughput satellites in geostationary orbits for national and regional applications, and the recent statements from the ITU regarding resource sharing, we expect that geostationary satellites will continue to be an essential component for commercial, civil and national security applications for countries worldwide — even as more and more countries add NGSO components to their networks.
Viasat is working closely with both global LEO and regional GEO satellite operators to integrate hybrid multiorbit, multiband satellite communication networks to deliver the benefits of each for civil, commercial and national security missions consistent with globally shared and sustainable orbital resources.
No. Geostationary satellites will not become obsolete. But they will likely decrease in importance over time, and there will be less motivation to upgrade these satellites as they become confined to a backup role. The clear success of Starlink despite early skepticism has highlighted the potential advantages of using LEO once a satellite operator gains sufficient scale. The problem is that smaller operators (and even many of the major legacy GEO operators) don’t have the vast resources of companies like SpaceX and Amazon and so can’t keep up with the investment required to build and maintain a cutting-edge LEO network.
The Starlink constellation has grown rapidly since the first test satellites were launched in 2018, growing to more than 6,000 satellites today. As SpaceX maintains a regular schedule launching more and newer Starlink satellites, the company has rapidly grown its capacity throughput: Starlink has demonstrated it can deliver massively more capacity than existing GEO satellite-based communications services and with lower latency. Viasat recently posted on X that its in-flight connectivity service can handle connecting 50 million passengers a month. Assuming (optimistically) 1 gigabyte per passenger, that works out to around 50 petabytes of data each month — roughly what Starlink carries each day.
Take terrestrial consumer broadband as another example. The average U.S. Viasat customer pays around $115 per month for about 50 GB of data, while the average U.S. Starlink customer pays $120 per month for 300 to 400 GB of data. And we see that same higher capacity in other parts of Starlink’s business, such as maritime connectivity, with the typical customer using five times as much data over Starlink compared with GEO services such Inmarsat.
Across all its services, Starlink provides data with much lower latency. This may not matter for all customers. But it’s worth noting that over the past 20 years of mostly failed attempts at bringing in-flight connectivity to the masses, passengers have repeatedly complained that the experience was like drinking through too small a straw.
But while Starlink is the market pioneer, other players are trying to chase Starlink’s success in LEO. OneWeb already offers internet broadband services, though its 600-satellite constellation cannot compete with Starlink in capacity, and even the second-generation system planned for the early 2030s is unlikely to close this gap. The vast resources that Amazon has committed to its Project Kuiper satellite broadband constellation makes it the most plausible competitor to Starlink, and it will begin launching satellites aboard ULA, Arianespace and Blue Origin launch vehicles in early 2025. Given that GEO operators are struggling to get their newest generations of GEO satellites in orbit on schedule — and they are further hampered by not simultaneously operating a launch service like SpaceX — they will face stiff competition from the rapidly growing LEO services.
But just because Starlink offers greater capacity doesn’t mean every customer will abandon their GEO vendors overnight, or even ever, if they don’t need much data and GEO service is adequate for their needs.
For example, credit card verification at gas stations has continued to rely on older GEO satellites for the past couple of decades. But when it comes to in-flight connectivity, airlines will have to decide if they can afford to offer the service to their customers for free to distinguish themselves from the competition. Easy to use, high-quality, free service is what Starlink is emphasizing, but it comes at a significant cost to the airline. So the billion-dollar question in the industry right now is whether passengers will choose to fly a particular airline because it has Starlink Wi-Fi rather than Viasat Wi-Fi, so airlines will be forced to upgrade their service to retain market share. We haven’t seen much evidence of that yet. But it is still early days, and some airlines like United Airlines clearly hope that Starlink will provide them with a significant level of differentiation in the marketplace.
No. GEO will not become entirely obsolete. I do believe that LEO, given the performance advantages it offers in terms of lower latency and a much more distributed, resilient network, will become the preferred architecture for broadband connectivity requirements.
Starlink has certainly validated how powerful LEO is, and we believe our Telesat Lightspeed broadband internet network will be a competitive, transformational alternative for enterprise, telecom, government, aviation and maritime customers. Unlike LEO networks that were designed primarily for best-effort consumer broadband connectivity, the Telesat Lightspeed has unique capabilities for enterprise customers, including committed information rates and service level agreements, multi-Gbps [gigabytes per second] data links, and complete control on the services delivered to each of its sites. Telesat is now in the engineering and manufacturing phase of the project, and we expect to begin launching our LEO satellites in mid-2026 and begin beta testing shortly thereafter, with global services commencing in late 2027.
There likely will be some user segments that will value the benefits of a multiorbit offering of GEO and LEO, including government users who prize network resiliency. The U.S. Department of Defense, for instance, is moving away from using a handful of easily targeted GEO satellites for communications, missile tracking and other services. Its Proliferated Warfighter Space Architecture network will rely on a new constellation of many small military satellites in LEO while also integrating existing commercial LEO constellations as part of the architecture. The Telesat Lightspeed network was designed from inception to achieve a high cybersecurity rating under the U.S. Space Force Infrastructure Asset Pre-Approval program and to be interoperable via optical laser links under the Space Development Agency’s standards. GEO will continue to be the most efficient distribution option for direct-to-home broadcast networks and television distribution, albeit this is no longer a growth segment in most parts of the world due to cord-cutting. Certain networks in tropical or harsh weather environments will likely maintain C-band links for network resiliency, whereas UHF and X-band payloads in GEO will remain valuable to defense users.
I believe GEO will have a role to play in the future — but most broadband connectivity requirements will be better served by low-latency LEO networks.
The customer community — whether it’s consumers or enterprises — see huge performance benefits in LEO services. Recently, United Airlines and Air France announced they are transitioning from their GEO networks to using Starlink for in-flight Wi-Fi connectivity. And Starlink has made significant inroads in the maritime and consumer broadband connectivity markets, with over 4 million subscribers.
Many GEO satellite operators and service providers are partnering with LEO operators to complement their existing satellite infrastructure. Viasat, for instance, recently announced it is in discussions with Telesat for LEO capacity and has also partnered with Eutelsat OneWeb for maritime multiorbit connectivity offerings.
No. GEO satellites are not currently competitive with LEO constellations, but I don’t think it’s a problem intrinsic to GEO. It’s a problem with the way we’re using GEO.
Spending $700 million to build and launch a GEO satellite is no longer competitive with new LEO constellations. The first step to innovating around that challenge is to bring down the cost of GEO satellites, and that previously has only come from shrinking them. But smaller satellites present a problem: less power.
Power determines throughput, and smaller GEO satellites today generate less than 5 kilowatts of power compared to the 20 kilowatts of traditional exquisite GEO satellites. If you pay $300 million for a communications satellite that only gives you 5 kilowatts of throughput, the business case doesn’t close on a dollar-per-megabit basis.
We founded K2 Space in 2022 with a goal of reducing the cost of building GEO satellites with more power, taking advantage of what we call “the new era of mass abundance.” We’ve watched the success of launch vehicles like Falcon 9 push launch costs down, and Starship and Blue Origin’s New Glenn are set to reduce the cost of mass to orbit even more. So we’ve designed our Mega Class satellite with a 20-kilowatt bus. At $15 million and the ability to stack three GTO (geostationary transfer orbit) satellites in a launch vehicle, we think the unit economics of GEO start to make sense again. Our next model, the Giga Class, will be a 100-kilowatt bus. But we’re not building big satellites for GEO just because new launch vehicles make it possible. We believe there are important advantages to GEO.
First, unlike SpaceX and Kuiper, most operators don’t own their own launch vehicle. We think those operators will find it almost impossible to deploy a LEO constellation at scale and make the business case close.
Second, you can more quickly deploy greater capacity in GEO to target a service area. Doubling the capacity of my LEO constellation over a region requires doubling the size of the entire constellation, requiring multiple launches. To double my capacity over a specific market at GEO, I only need to put up a second satellite in that GEO slot.
And the higher you go, the fewer satellites you need for global coverage and the fewer launches you need. With less expensive, more capable satellites, we envision a multiorbit future with layers in LEO, MEO and GEO, all with their own use cases and customer segments.
No. GEO satellites will not become obsolete, but GEO communications have probably passed their heyday.
I have to admit that when SpaceX and OneWeb announced in 2015 that they were building LEO constellations, I was skeptical. Almost a decade later, my conclusions have almost flipped entirely. It now appears that for many or most applications, a sufficiently scaled LEO constellation can provide better service across almost every attribute compared to GEO constellations.
Take maritime communications. A standard GEO antenna weighs more than 200 kilograms and costs $25,000 or more. Alternatively, a sailor can carry a Starlink antenna aboard under their arm for $2,500. The GEO antenna will deliver data rates of 20 megabits per second at best, while Starlink provides 100 to 200 Mbps, and with 50 milliseconds latency compared to the 800 ms for a GEO satellite system.
Putting performance aside, the GEO industry also faces challenges in terms of capacity and timing. It took Viasat seven years and reportedly about $700 million to design and build its latest satellite, the ViaSat-3, and so far, only one of the planned three satellites is in orbit — the satellite serving North America launched in the spring of 2024 and experienced an antenna issue that reduced its capacity by 90%.
Companies such as California-based Astranis and Switzerland’s SWISSto12 have begun developing smaller GEO satellite alternatives to the multiton warhorses common in GEO. The idea is to fly a relatively larger constellation of cheaper, smaller GEO satellites to save costs and speed production. But LEO constellations like Starlink have a significant head start. Even if GEO was toe-to-toe competitive with LEO today, supply chain issues would make matching LEO constellations in capacity difficult.
But GEO satellites do offer some advantages over LEO. There are areas of the globe, such as China, where access to the internet is tightly controlled, and consumers are not permitted to access Starlink. That leaves GEO services — mostly Chinese owned — as an alternative. And by virtue of their high altitude, GEO satellites are less affected by space debris or geomagnetic storms, like the one that pulled more than 40 newly launched Starlink satellites from orbit in 2022.
Large maritime vessels still retain their GEO antennas, even as they shift most of their operation requirements over to the LEO system. The iconic “R2D2” dome isn’t going away anytime soon, but its role is quickly shifting to that of a backup system.
Related Topics
Communications satellites
