As a wave of plans for low Earth orbit (LEO) broadband constellations enters the market, a 63-year-old former NASA Branch Chief and Google executive is betting on narrow beams, higher frequencies and private-network architecture to meet growing demand from government and enterprise users.
Founded in 2023 by Milo Medin, who earlier built terrestrial broadband provider Excite@Home during the early days of the commercial internet, Logos Space Services recently secured FCC approval for a constellation of up to 4,178 broadband satellites. The network would operate primarily in K-, Q- and V-band spectrum, higher frequencies than current LEO broadband systems that can improve capacity and interference resistance but also pose engineering challenges.
Already this year, Blue Origin has unveiled its TeraWave network and Open Cosmos has outlined plans for a sovereign European offering, alongside other entrants targeting government and enterprise markets once seen as less crowded than consumer broadband.
For Medin, the timing reflects both a changing threat environment and how far satellite communications have evolved over the past three decades. Logos is backed by serial entrepreneur Thomas Tull, whose dual-use technology investment firm U.S. Innovative Technologies led a $50 million Series A funding round for the venture last year.
In this interview, which has been edited for clarity and length, Medin discusses Logos’ strategy for private-network connectivity, how the venture is addressing emerging electronic warfare threats and what comes next after regulatory approval.
SpaceNews: What drew you to building a satellite network at this moment in the industry’s evolution?
Milo Medin: There are two factors driving why this is an interesting and necessary product space. First, a lot of the exploitations, breaches and hacking activity going on in the public internet has extended from the core operating systems to virtual private network (VPN) concentrators, firewalls and other infrastructure.
Most enterprises are not interested in talking to Russian trolls and Nigerian princes. They want to get data traffic from A to B, and they don’t necessarily want complete connectivity along the way. We have private networks in the wireline world, and I think there is a need for something that delivers the same kind of pervasive connectivity a space-based system can do, without necessarily having to go through infrastructure that’s designed for the internet.
The second driver is jamming and electronic warfare. Space systems have been designed for a relatively peaceful environment, where interference is generally accidental and not deliberate. That is not going to be the norm. Jamming is not just a vulnerability for geostationary Earth orbit (GEO) but for low Earth orbit as well.
You have to design your communications system, waveforms and spectrum approach to be resilient to jamming, and that’s very difficult to back into a network architecture later. You have to design for it at the beginning.
This problem space was posed to me by Thomas Tull, our chairman and the primary sponsor of the company today. He believed this was a big problem and asked me to look into it. That’s how we started, so he really deserves to be labeled as a cofounder.
It’s the flip side to space networks becoming more prevalent and critical.
Yes. I’m old. I remember the early days of the internet, when we didn’t have nation-state actors and ransomware and all the rest of those things. It was a peaceful environment filled with academics. When break-ins started happening and security holes were exploited, no cost was imposed on the people doing it, so you got more and more of it.
I think the same is happening now. Unless cost is imposed, you’re not going to have a negative feedback cycle. Consequently, more and more dollars are being put into jamming. That’s why I think we’re on a trajectory where contested spectrum will be the norm, and our systems need to be designed to deal with that. It’s a nontrivial challenge.
The FCC approved Logos to operate in K-, Q- and V-band spectrum to help address that, though some segments were deferred or require additional coordination. Did they change anything about your near-term plans?
They really haven’t. The FCC did a fine job of trying to move through this aggressively. It was a complicated spectrum process. We started working on federal coordination before we went to the FCC and we’re still working through the rest of it.
We believe our architecture, using super-narrow beams, allows us to coexist with systems in ways that non-narrow-beam systems would have difficulty with. Because we’re not trying to connect millions of users, we can put our energy in very narrow pockets, and that makes it easier for us to coexist with existing systems.
We also tend to come down at higher angles and not shallow ones. That helps not just with coexistence and managing interference, but also with resiliency to jamming. At Google, I ran the Citizens Broadband Radio Service (CBRS) effort, where we similarly worked collaboratively.
What are the immediate next steps now that the constellation has regulatory approval?
We’re working through launch provider agreements as well as bus selection. We are probably the first large constellation that believes we don’t have to vertically integrate the bus.
Later this year, we expect to down-select to a smaller set of bus candidates and run a full request for proposals (RFP). We hope to have a bus award done by summer, or even before then.
We’re designing the payload and it will be integrated at the bus manufacturer’s facility. In a flatsat, it’s not like the bus and payload just clip together. It’s a very integrated design, so managing that complexity and making sure it works in production is important.
We’re not building prototype satellites. We plan to go into low-rate initial production, work through the production kinks and integration, and then go to serial production to build the constellation in a reasonable timeframe. A lot of the work this year is about how our payload gets integrated, how we do flight operations and how we line up the supply chain.
When could we realistically see the first spacecraft deployed and initial services?
I would love to see us begin launching spacecraft at the very end of 2028, but until we have our launch and bus contracts done, we’re not going to have a firm date.
The initial M1 shell is for providing service in the ± 32-degree latitudes and our hope is to have all 325 satellites launched within about six months. That depends on serial production and launcher capacity. So our goal is to have M1 operational in the 2029 timeframe. I’m hoping sooner in 2029 than later, but all of this is pending launch and bus contracts.
There were initial plans to begin launching the satellites by the end of 2027.
Based on what we’re seeing in terms of launch capacity and bus maturity, I think 2028 is the timeframe right now. We might be pleasantly surprised, but generally I prefer not to depend on pleasant surprises.
Several networks aimed at government and enterprise markets are also coming around the same time. What does Logos need to do differently to survive that competition?
Part of it goes back to using optical inter-satellite links (ISLs) that enable us to go up, over and down without having to have terrestrial infrastructure in the area. Part of it is also how we think about spectrum, waveforms, apertures and antennas.
It’s hard for me to explain differences with systems that are very early in the design phase, or haven’t gone through the FCC process. One thing I do know is that physics always gets the last vote.
One advantage we have with this spectrum is that we don’t have GEOs above us. If you’re operating in conventional Ka or Ku, you have to protect the GEOs in the equatorial belt from harmful interference. That means you’re not talking straight up. Satellites like Starlink have to talk at an angle, and that can create availability issues and make you more vulnerable to jamming from terrestrial sources.
By talking straight up — which we can do because there are no GEOs in the uplink band — we avoid that. If a terrestrial jammer is coming in horizontally, it has to go through clutter, and by the time it reaches the aperture it’s substantially attenuated. That’s one of the reasons we can build a system that is resilient to that kind of jamming, and it goes back to why we picked this set of uplink bands instead of commercial Ku or Ka. And it would drive adversaries to have to spend resources on jammers for new bands as well.
Higher-frequency spectrum also brings challenges like rain fade and stricter pointing and terminal requirements. How are you mitigating the downsides in the network and terminal design?
We’ve done a lot of analysis based on scientific data and on what operators are doing there today. I don’t think it’s as bad as people think. TeraWave picked the same downlink band, for that matter.
Part of it is that you need enough dynamic range — the ability to change signal power from the spacecraft — so that you can deal with rain fade. That’s part of the rules we negotiated with the FCC, in terms of how we can manage transmit power.
One of the nice things about the narrow beam is that if I have to increase power from the spacecraft to get through weather, I’m not increasing power everywhere. I’m increasing power only in that narrow area affected by the weather.
We also applied for 17 gigahertz spectrum, and that can help in areas with really heavy weather. We think we’ve got very good availability at Q- and V-band with the terminals and spacecraft apertures we’ve designed, but the ability to use 17 gigahertz as a backup in the heaviest-weather areas will help us maintain contact and performance there.
We’ve seen many constellation plans come and go over the past decade. What are the biggest pitfalls you’ve seen derail megaconstellations, and how are you designing Logos’ financing, manufacturing and deployment approach to avoid them?
One of the benefits of doing it now, as opposed to being first out of the chute, is that you learn from other people’s experiences. We’ve got folks here who worked at SpaceX in the early days and on the Amazon Leo program, so we’ve got a lot of experience in terms of what worked and what didn’t.
The number one thing I worry about is integration of all the components and the supply chain. Optical ISLs, phased-array antennas — all of that has to come together in a timeframe that not only lets you build the first few spacecraft, but enables you to ramp to serial production so that the economic model works. If it takes you five years to deploy the first shell, that’s a problem.
The first shell is only 325 satellites. You can look at that and say, that’s not very much, or you can say, holy cow, that’s bigger than any of the U.S. Space Development Agency (SDA) tranches. It depends on your perspective.
On the financing side, we will do it in tranches as we make progress against internal milestones. Right now, we see a lot of demand for our product. After working on large infrastructure projects before, I’ll say this: if there’s customer demand for what you want, you can get financing to go build it. If people don’t want to buy what you’re building, that’s a big problem.
If we execute well on building what we want to build, and the demand is still there, I don’t think financing is the main problem. It’s really execution against that agenda that will drive success.
During your time at Google, you helped spin out laser terminal and network orchestration platform Aalyria, which recently achieved unicorn status after raising $100 million. As constellations add optical links and become more software-defined, how does real-time network orchestration change how these networks operate, and where might the industry be underestimating the complexity?
In the GEO world, everybody used the same modulation and roughly the same power levels, and you could move a terminal from one GEO to another. Then OneWeb and Starlink came along in LEO, and it was like, okay, it’s all Ku, so we’ll have one terminal that can talk to GEO or LEO and work that out. At some point, that process becomes very painful to accommodate.
What you really want is not one platform but multiple platforms that you integrate. That’s how the internet works. Rather than trying to make one terminal that talks to everything, or one satcom network that does everything for everybody, you enable connectivity between multiple different systems and terminals.
The difference in what Aalyria is trying to enable is not just connectivity, but the ability to manage connectivity with quality of service, bandwidth control and other performance expectations.
If they can pull that off, that enables a space-based set of systems to do what the terrestrial systems have not done in the internet. It would also mean you’re not subject to just one system’s weaknesses. You can take advantage of different LEO systems, maybe even GEOs, and stitch together capabilities depending on what you need.
That moves away from the idea that one thing is going to do everything for everybody. That’s a different approach to satellite communications. It’s very internet-like — lots of different networks that you can interconnect and choose from — but it’s also trying to manage performance across that environment while things are moving around. It’s a very dynamic environment.
We’re also seeing more countries push for sovereign connectivity, and some newer entrants are positioning themselves explicitly as domestic alternatives. How do you think about serving non-U.S. government and enterprise customers in that environment?
If you are delivering internet, you’re part of one routing system. With a private network, we have the ability to create virtual partitions in the system, with chunks of capacity allocated differently.
I think there’s a question about what it really means to have sovereign capacity. If everybody has to travel the same routing path, that’s a challenge. If you can create a private network for some foreign governments and users, that may give them the kind of differentiation they want.
Ultimately, it depends on what your goals are. If you don’t want to depend on anything American at all, that’s a tough row to hoe, because you’re going to want your system to operate over the United States, too. But if the goal is differentiation in performance, security, distribution of service or who sells the product, we may be able to accommodate that.
And it’s a geopolitical situation that appears to be constantly shifting.
Ultimately, we want to build a product that meets security and resilience goals for enterprise and government users, and if that can meet the goals of some of these foreign nations, we’d love to serve them.
What do you think the next major technological shift for satellites will be?
Use of higher frequencies, maybe nontraditional spectrum, is definitely one of the shifts we see going on. We’re part of that. The radios, antennas and capabilities we can use to communicate at higher frequencies are easier to do today than they were only a few years ago.
Obviously, the ability to communicate directly with handsets is a big change, and we’re seeing a lot of energy around that. I wonder about whether the economics work, but as an engineer it’s impressive to see that ability.
I also think we’ll see a move to bigger satellites with more capability on them, along with the ability to use higher frequencies to increase data rates and deliver new kinds of services.
And spacecraft are becoming platforms. We think about them not just as a network, but as a platform where you can do other things on them. As you get proliferated constellations, you can get proliferated sensors and relay capabilities.
That’s going to make space a much more capable environment for delivering services — not necessarily competing directly with fiber, but delivering fiber-like services and capabilities terrestrial networks can’t.
