Sky High Promises: The Limits of Starlink - Episode 657 of the Community Broadband Bits Podcast

Christopher Mitchell (00:12)
Welcome to another episode of the Community Broadband Bits Podcast. I'm Christopher Mitchell at the Institute for Local Self-Reliance. I'm in St. Paul, Minnesota and for possibly, I don't know if we've done this before. We might have done it around some of the BEAD stuff but we've got a back to backer. We got Sascha Meinrath back again from the ⁓

though illustrious, I don't if illustrious is a word that's relevant here, but Penn State University where the land grant university that served my community when I was growing up. So I have a lot of affinity for it, but he's the Palmer Chair in Telecommunications there. Welcome back, Sascha

Sascha Meinrath (00:52)
Pleasure to be here. Of course, I'm coming to you from Canada today. So just switching it up

Christopher Mitchell (00:57)
Yeah. Yeah, I'm looking

for that accent. So you're going to have to work on that. You're also the founder of X-Lab And we're going to be talking about some cool research that you've done. I don't know which hat you were wearing when you did it. We maybe won't touch on that. But we're going to talk about that. And in order to have a more full conversation about it, because I thought that people get bored of just hearing you and me continue on after our fun session two weeks ago.

We got Sean Gonsalves, who is Associate Director for Communications at the Community Broadband Networks at the Institute for Local Self-Reliance. Hey Sean, welcome back.

Sean Gonsalves (01:33)
All right, thank you.

Christopher Mitchell (01:34)
And then I cajoled Ry into coming out of the Fortress of Solitude there in Mankato to ⁓ join us as a live guest rather than producer. Thank you, Ry. How's it going?

Ry Marcattilio (01:44)
going well. Good to be here, Chris. Good to see you all.

Christopher Mitchell (01:46)
And you are the Associate Director of Research here at the Institute for Local Stuff Reliance and the Community Broadband Networks Program. Cool. So Sascha, you've released some cool research that you've been working on for a while with others. I know I'm sure you can credit them regarding Starlink. So today we're going to be talking about Starlink with some empirical results and just filling in some gaps of things in terms of what we're seeing as it is expanding rapidly.

⁓ And I think we should be able to cover this from different perspectives, not just sort of like one of horrified, a horrified reaction to the idea that this should be our sole way of connecting rural America that some people seem to support. So Sascha, what did you release?

Sascha Meinrath (02:27)
Well, so there's long been this question, and it is an open question. What is the capacity of satellite connectivity to be a broadband solution to bridge the digital divide to serve those, and there are millions, who are on an underserved across the United States? And the problem one runs into is the companies that provide this and the companies that have traditionally provided this are super secretive about what they're

actual throughput capacities are. So we scoured the globe and spoke with lots of research labs and pulled together kind of the best available data. And I'm hedging here because I want to acknowledge the fact it's an exploratory analysis based on this best available data. And if Starlink would just release some better technical specifications, we might narrow in on what the actual capacities are. But the take-home message

is that we were shocked by the limitations to Starlink capacity and what it really drives home is the need to do due diligence on, in essence, where and when you can use Starlink without causing, in essence, catastrophic degradation of service that brings it below the threshold to meet criteria as broadband. Which is to say, states and the federal government are in danger

if they don't run these kinds of due diligence analyses of illegally, and I say that because the statute requires broadband service provisioning, of illegally subsidizing sub-broadband service for potentially millions of Americans. And that is a real problem. It's a problem of waste, fraud, and abuse. It's a problem of sort of like allowing the PR and the vaporware administrations of Starlink

to guide as opposed to math, science, and physics. But it also is a real problem because those on the wrong end of receiving this service will get a substandard service that is simultaneously exorbitantly priced vis-a-vis fiber connectivity. So it's double disadvantage for those that are subjected to an oversaturated Starlink network. Our analysis demonstrates using math.

why this is a problem, and I'm waiting. I Washington Post covered it. We reached out to Starlink to try to get there. Like, we're waiting for some sort of response. Either tell us if we're wrong or acknowledge that we're correct. Thus far, Starlink and NTIA have been completely mucked on this foundational problem.

Christopher Mitchell (05:03)
And so what you're saying is that like, if I could sum it up in a slightly different way, and you can tell me where I get this wrong, because this is helpful to identify gaps in our communication, we don't really know what Starlink is capable of. And what compounds that to some extent is also that it's changing, right? On a regular basis, depending on what orbits are used, on how many satellites are up there, how many are degrading, how many are new or launched.

To some extent, one might think we could go by what Starlink says. However, and you didn't touch on this, but I think this is a wrinkle. However, most of what we learned about Starlink, I think, comes from Elon Musk, a person who has just said fantastical things without any evidence. He's made predictions for more than a decade in terms of like full self driving for the Teslas that have just been ludicrous, you know, to the point at which he said it would be fully available on the order of like eight years ago, I believe, and every year since then.

And it's never happened. And so I'm not saying this to say that he's a bad guy or that like, you or that, you know, we can't trust our link. But the simple fact is, is that the company's spokesperson is known to have lied on many occasions. And so I feel like you're trying to say, okay, let's get beyond sound bites and let's get into the tech and figure out what actually will work because we're about to turn tens of thousands, hundreds of thousands, perhaps millions of people onto this.

And that's not just across the United States. That might just be in like a region of the United States.

Sascha Meinrath (06:27)
Yeah, it generally can curve, but we do actually have some parameters, which is to say there are certain physics-based limitations. We know the channelization of these satellites. We know how much spectrum they're using, for example, the uplink and downlink capacity of these satellites. And there's this thing, Shannon's Law, like we know the theoretical maximal amount of throughput you can put on there. And so those are the kind of things that then guide our analysis.

We're using physics as the foundation and in that we get sort of these parameters that say, look, the maximal amount we can fathom is this. And when you start working back from there, you realize like that's not a lot. That because of the way Starlink is architected.

Christopher Mitchell (07:07)
Mm-hmm.

Sascha Meinrath (07:11)
and we were able to reverse engineer that from a number of their FCC filings where they are required to disclose certain elements of this. There's just fundamental limits to the architecture of their network. And we explicate those and we make the model very, very transparent because we want people to bang on it, flip around some of the variables and try different parameters. But it does give us at least this kind of ceiling of how much capacity it can handle.

Christopher Mitchell (07:40)
I'm about to bring in Ry and Sean, but before I do that, can you tell us to what extent your use of ⁓ actual speed tests and such data collection have informed this analysis as opposed to theoretical limits and things like that?

Sascha Meinrath (07:54)
Yeah, so we did a bunch of analyses a number of years ago and continuing to this day where we've collected millions of speed tests and so we have a pretty good grounding in reality that pointed to something being wrong. I had one of the first Starlink terminals, Dishies, in my backyard, in my region because we wanted to actually test what was happening there and there's a number of phenomena there from latency spikes to just degradation of service as more and more people in the region.

and get on that we documented. And we also have things like Ucla, which has released publicly this massive data set that's showing that 17%, that's 1.7 % of Americans, are actually getting broadband via Starlink. That's not a good, like if 80 % of folks are not getting broadband, that's not a good starting point for this. This is before you start saturating the network with a whole bunch of new users.

So it's only going to get worse from here. So the speed test results and the physics align. They triangulate on a phenomenon, which is that this network is already under strain. It's not able to consistently provide broadband services. And that is going to get worse.

Christopher Mitchell (09:08)
Right, and I just wanted to note one of the things that I don't know if everyone knows this, but you're a pretty technical person, which is helpful because there's people that wade into this and they don't understand the tech. like to say that Starlink is not providing broadband is an interesting assertion because it actually comes into a time question, right? At any given instance, what is Starlink capable of providing based on what do you have in line of sight in terms of satellites? How much under capacity or under strain are those?

or those individual satellites going to a ground station somewhere, there's a bunch of different stuff to come into effect. But what you're saying is that of the individual speed tests that people have been doing on the regular, a strong majority of them, on 83 % of them, if I did that, if I remember correctly, you 17%, are not getting 100 megabits down and 20 megabits up, which is the current federal definition of broadband.

Sascha Meinrath (09:58)
That's correct, and that's before we get into things like latency, which again, the broadband has a third component, uplink, downlink, and latency, the time it takes a packet to traverse a network has to also be...

Christopher Mitchell (10:10)
Right. can for people who are confused by that.

This is like if you remember, I always think, I don't know, maybe the people on this, on listening to the show remember calling across the continents in the 80s or the 90s or even to the 2000s. I remember I was overseas in the year 2000, did a lot of those calls and you'd consistently have collisions when you try to talk because the other person, you weren't sure if they were done talking or not. Because from the moment that they started talking to when you hear their voice on the other end of the planet,

⁓ it could be a significant amount of time. And so that's latency. Go ahead, Sascha.

Sascha Meinrath (10:40)
That's right.

And if anyone's ever had a phone call or zoom call where things have pixelated or frozen, that's usually also latency. So we're experiencing this all the time. Anything beyond around 50 to 100 milliseconds is bad. Anything beyond 100 milliseconds is not considered broadband. And what we've consistently seen spanning millions of tests all across the United States over four years are latency spikes in the 600 millisecond and up range, which is

Christopher Mitchell (11:09)
spikes.

Sascha Meinrath (11:10)
functional. Yes.

Christopher Mitchell (11:10)
Right. I did just see a separate report, I think. I don't think it was yours. And it suggested that commonly Starlink, I think on the order of a median latency, was on the order of like 50 to 60 milliseconds, which it noted was much worse than the 20 to 30 that we experience on mobile networks. And I believe on fiber optic networks, we're in the single digits ideally, but sometimes in the 10 to 20 range. Is that all correct?

Sascha Meinrath (11:37)
That sounds exactly right. And it's also the case that what we're seeing is handoffs between satellites causing consistently, and I mean like every three to 15 minutes, a latency spike in the 60 to 100 millisecond range or higher. That is part of the architecture. You can't route around that unless you were to do something completely different than LEO low earth orbit satellites.

That is a fundamental problem that cannot be fixed as long as you've got satellites whizzing across the sky and having to hand off the connection to other satellites.

That is, as far as I can tell, completely externalized by NTIA in terms of the quality of service problem that we're also talking about. So it's speed, yes, and it's latency. We looked at speed in our analysis, but we also want to just put a little pin in the fact that latency is also an issue on this infrastructure.

Christopher Mitchell (12:36)
Now,

Ry, I'm curious, I don't know if you remember this, I seem to recall, NTIA basically said everything underneath 100 milliseconds of latency is the same. It basically sets a threshold of like, can't be over 100. And I remember being a little bit horrified at the idea that we should treat eight or 10 milliseconds of latency the same as 99.

Ry Marcattilio (12:55)
Yeah, that's my memory too. And to be clear, you know, you might not notice higher latency levels if you're watching Netflix or doing like basic web stuff, although websites will take longer to load. But if you're playing a video game or something like that, you can absolutely feel the difference between eight and 25. And once you get into the, you know, 40, 50, 60 range, life is a little bit different on that side of the computer screen. so there's a bit when you get to 100, it's a radically different world.

Christopher Mitchell (13:21)
So Ry, what did you want to bring up here as we're grilling Sascha on this report and talking more generally about Starlink?

Ry Marcattilio (13:28)
You know,

as Sascha was talking, I was just reflecting on the fact that during the whole of the Biden administration and all the BEAD work we did over the last three years before that, I'm just struck by how little we know about how ⁓ Starlink operates and how we have to reverse engineer this from the back end, how much they remain a black box to ⁓ even the most technically minded people and ⁓ the disservice that that's doing to us today. I wonder

if we should have collectively done more work from 2021 to 2020, the end of 2024, you know, not only doing some more defensive work to protect from an eventuality where we live in a world where it looks like LEO is gonna get a bunch of this BEAD money.

Christopher Mitchell (14:07)
Sascha, what would that have looked like? I I think you've been thinking about this along all those terms. Would it have looked like more than your report or what should we have been doing along? And that wouldn't form what we should be doing now, I think, as we move forward. And would hope that in the future, we'll have an administration that actually cares about delivering high quality access to everyone and making sure that they're actually getting a good service.

Sascha Meinrath (14:32)
Yeah, well there's two ways to do this, right? So one is you do the engineering and you empirically document a priori beforehand that a solution can meet the need that you're hiring or paying for that solution to meet. The second would be like substantial and meaningful repercussions, usually in the form of like financial repercussions, if you fail to deliver what you've been paid to deliver. We don't do either of those, so you basically

Christopher Mitchell (14:58)
You know, I'm reminded of James Earl

Sascha Meinrath (15:01)
you

Christopher Mitchell (15:01)
Jones at the end of Sneakers when he says, we are the United States government. We don't do that sort of thing.

Sascha Meinrath (15:08)
he was so young and naive then, yes.

Christopher Mitchell (15:10)
You

Sascha Meinrath (15:11)
No, but in essence we don't have meaningful accountability and so all we're left with is an assumption that this thing is going to deliver. Fiber is pretty tried and true. We can demonstrate via pretty much any manner of analysis that it's got the throughput capacity. Starlink and LEO is not tried and true. In fact, consistently for decades it is underperformed. So all of the historical analysis is like this thing doesn't do what people are claiming.

And the only claims we have that it is able to do this are in the form of banter from a serial liar. So yes, I think one would be wise to trust but verify these claims, and we don't do that either.

Christopher Mitchell (15:57)
Do have

a sense that the Department of Defense has dug into some of this? mean, they're using this. I would expect that they would have a high standard. And I know that you have different contacts within the government.

Sascha Meinrath (16:09)
Yeah, so I...

It's not confidential per se, but the US military did reach out to us after we did our analysis and was like, thank you for doing this. This has raised profound concerns on our end of things. So yes, we know they're looking at it. We know they are also desperate for independent verification on the capabilities of this system. As you can imagine for them, it's like it really is a life and death situation if their communications gets funky.

For me,

the fact that they haven't done that yet and they're looking to like, know, whippersnappers like me to do that is a little terrifying, but also shows you the extent by which proper due diligence has been ignored in this space, much to the detriment of the lives and livelihoods of U.S. military and rural residents.

Christopher Mitchell (17:00)
Sean, what do you think about it? I you've been talking to different states, so you can't share everything that you've learned necessarily right now, but what's on your mind as we talk about this?

Sean Gonsalves (17:09)
Well, the first thing that's on my mind is just kind of how frustrating this is in that, know, LEO was always gonna be a part of the mix before this NTIA sort of changed the rules and deprioritized fiber. know, states like Maine, I think New Mexico, I mean, there were several states that were already gonna use Starlink as part of the mix, but now here we...

Christopher Mitchell (17:30)
Right, Sean,

you're responding, I think, to this claim sometimes that like that that we're anti-LEO. And I don't think that's accurate.

Sean Gonsalves (17:36)
Right.

Exactly. mean, so all along we and I think even, you know, sort of the whole BEAD framework was all we saw a LEO as being part of the mix. But now we're in this period where it looks like this is just a play to just sort of, you know, get Starlink some extra subsidies. But the thing in particular in your analysis, Sascha, is...

because we're talking about this scale of the digital divide and how many households, particularly in rural America, don't have access to broadband. Your analysis, I think, found that no more than six households per square mile within a beam could be served. And so that to me is sort of like a really important number. I mean, granted with all the caveats of the assumptions and as Ry said, this black box of Starlink, but...

But that is something that is a number that people should be thinking about because I think there is this popular perception that somehow Starlink could just take care of all of rural America.

Sascha Meinrath (18:39)
That's right. So as soon as you hit about seven served households, and that's new and legacy households per square mile within the service area of one of these beams that comes down from the satellites about a 63 odd depending Square mile area Then you get a saturation where the uplink capacity these satellites gets completely degraded and that's not a lot of people right which is to say like you'll hit that and

Christopher Mitchell (19:03)
No, mean, you don't,

Sascha Meinrath (19:05)
a lot of areas.

Christopher Mitchell (19:07)
right, especially, I mean, if people are aware, I mean, I would say in general, west of the Mississippi, you probably won't hit that, west of the Missouri, you probably won't hit that, east of the Mississippi, I mean, you think of Iowa as being low density, but actually, if you check it out, you'll find that there's a lot of people where we have that level of density.

Sascha Meinrath (19:25)
Correct. And that's assuming also that you're not eating up things with something called routing overhead, just the channel information of like to coordinate amongst all these different devices and the satellites. You need space for that. ⁓

Christopher Mitchell (19:38)
But Sascha, when I've

talked with people about this, and you can come back to routing overhead if you want, sorry. The beams, my understanding is that a spot in Iowa could have beams from different satellites on it. And so at that point, are you able to do 12 or are you saying that there's still a limitation in your ability to serve homes within that given square mile?

Sascha Meinrath (20:01)
Yeah, so this is where you get into really complex mathematics around self-interference, hidden node problems, and other issues within wireless communications in particular. I happen to have a goodly amount of expertise. My teams over the years have actually built protocols to solve this exact problem. It is unique. You need a lot of calculus. But the long story short is you don't necessarily have this additive function.

in terms of your throughput. You add more beams in, you end up with this asymptotic function of overall network capacity in a geographic area. And you might be able to squeeze a little bit more capacity, assuming that you've got nodes that can't see one satellite but can see another and they're all in the same. But long story short, that then also fails to take into account that those other satellites, especially when you've got a national program that's providing LEO all over the place, they're tasked on their

spaces and don't have time to go and add capacity somewhere else. That in essence we're not talking about one dense little environment surrounded by Montana. We're talking about on a national scale this being rolled out because NTIA is privileging through their application scoring, a LEO all across the country. And so every satellite is very likely to be maximally

And so you are unlikely to have that extra capacity, those extra beams that you can put over in another location. The question in my mind is what happens when one state is thoughtful and does the analysis and is like, okay, we only want to have this many people in our state, but the state next to them goes full in and completely saturates the network and now all your satellites, it might be flying over

over

your state this second or busy over there and so you still get screwed. In essence it's a capacity over a region that's a common shared across all these states and so every state broadband office has to be like so even if we do the right thing here are our neighbors also doing the right thing because if one of them abuses this we're all screwed.

Sean Gonsalves (22:19)
So actually speaking to like state broadband offices, I was curious to know if you can share. Have you heard from state broadband offices? I mean, we've talked to a few and I think word is starting to get out there and some seem to have, they're looking at performance matrixes and sort of wanting LEOs providers to demonstrate capacity down the line.

assuming increased demand, we won't say which states, but I'm curious from you, have you been talking to state broadband offices and what's been the reception?

Sascha Meinrath (22:46)
So I've talked prior to this analysis being public with dozens of state broadband offices over the years. Post this, we're only beginning to have state offices ask some questions. And we're thankful for those state offices. I feel like the ones that are now like, ⁓ we need to address this are the wise ones. They're the ones that I have more faith in because it's a problem, a fundamental problem of physics. Like you can't rat around that no matter how many good intentions you might have.

But it is still few and far between that are doing this. It is a new wrench in the mix for a lot of them, right? Which is to say they're under a federal mandate from NTIA that they must provide service to the lowest bidder, but they're under a statutory mandate from Congress that says you must provide broadband.

And this is like the unstoppable force meeting an immovable object, a problem that now is on their radars and that they need to solve. So how they do that is complicated. The necessity of doing that, that's what we want to drive home. They must do this. Otherwise they are knowingly providing substandard non-broadband service to an unknown array of their population.

Sean Gonsalves (24:01)
Do you think there's any prospect that like maybe even some like big or any, not necessarily just big, but ISPs that might, you know, don't know, sue and say, look, we could have got some of that BEAD money, but it was awarded to Starlink and they can't even deliver. So, I mean, do you get any sense that that is a possibility or is that just a?

Sascha Meinrath (24:20)
Yeah, I mean I'm reminded of the fiasco for barrier free, right? This is an ISP that just claimed over six states to have universal gig symmetric fiber that didn't exist and eventually Yes, and eventually the FCC did find them this tiny amount to say like look you can't just make stuff up Well, the capacity here is made up LEO does not have the capacity to provide broadband on a national scale like it simply doesn't exist so to pretend that it can't

Christopher Mitchell (24:30)
It was remarkable growth.

Well, let me push you on that. Let me-

Sascha Meinrath (24:49)
is ridiculous. Go ahead.

Christopher Mitchell (24:50)
So, I mean,

so if I'm putting on the hat of people that are looking at Starlink as really pushing the boundary of what's possible and making impressive strides, I believe last year they claimed around this time, actually last year, about 1.4 million US customers. Now they're at 2 million, they say. That suggests a growth of on the order of 50%. While...

I believe your speed test would suggest that their quality hasn't really gotten much better, but it hasn't gotten worse as they've grown by 50%. I think that's the most relevant metric. They've also hit 6 million people worldwide, which is a growth of about 30 % over the end of the year 2024. That's not as relevant to me because if you have a bunch of countries where satellites are basically dormant going over them,

turning them on doesn't really stress the system significantly. So I'd focus on the US numbers. They've shown remarkable growth while being able to like continue to deliver a product that people more or less love.

Sascha Meinrath (25:52)
That's absolutely right. You know what else shows remarkable growth? Ponzi schemes. So like as long as you keep adding satellites and adding users, you can kind of keep keeping up with the demand until you can't add any more satellites, but you're still adding users, which is what NTIA is about to do. And who gets screwed? That last group that gets added in, in this case potentially millions of households. They are at the end of the Madoff chain.

Christopher Mitchell (25:57)
Yeah

Sascha Meinrath (26:20)
And they're the ones that get not the broadband service. And I see that as downright inevitable. And again, like, well, no, right? Like, maybe I have egg on my face. I have a pretty good track record of calling things as I see them and not being incorrect.

And I think this is going to be a case where the data is going to overwhelmingly show that if we just deploy LEO satellite broadband connectivity, that people are just not going to receive the statutory mandated minimal service that they are guaranteed by law under this program. And the authorities that are complicit in perpetuating that fraud, that's a real problem and they should be.

to account because their goodwill is coming up against physics and I'll take physics every day.

Christopher Mitchell (27:08)
Right?

Ry Marcattilio (27:10)
So I like that a lot, Sascha. We're all going to take physics whether we like it or not at the end of the day. I wanted to also play devil's advocate here and ask. So Tennessee just released its ⁓ most recent results of its benefit of the bargain round. if you look at the, they did it by the project areas by counties. And if you look at those counties, you could see that they got a lot of bidders, a wide variety of them, LEO and.

Christopher Mitchell (27:12)
Yeah, I mean, you don't have a choice. You're going to take physics whether you like it or not. Go ahead, Ray.

Ry Marcattilio (27:34)
fiber and some fixed wireless in there and some cable I think even. But in every single county, the LEO provider is like six to 10 times less, the bid is six to 10 times less than if there's a fiber wireline provider applying for that county. At the same time, Kuiper is underbidding Starlink it looks like in most of the instances that I saw. And so in a world now where Kuiper maybe grabs

you know, half of the BEAD money for LEO and Starlink grabs the other half of the LEO, the bucket of BEAD money that goes to LEO. To what extent do we think that, you know, relieves the congestion on Starlink, which we know, by the way, much, much more about than what Kyber's doing, to the extent that maybe doesn't look so bad in three or four or five years time.

Sascha Meinrath (28:16)
Yeah, so I want to touch on two things. One is the reason why satellite can be so cheap is because they've already got the capacity. They're over building themselves, which again is forbidden by law. So you've got this weird paradox currently where for the purposes of being eligible, of being a household that can receive broadband funding, you have to not already have service. These households already have service via the satellite. So NTIA is doing this duality where they're saying, okay,

like the satellite doesn't serve your home now, but you're going to get satellite service. And the satellite providers are like, well, that cost me like nothing. Like I actually don't have to do anything for that. So their over-ability, which again is a stricture that is forbidden by law. And again, it's a paradox of this entire thing. The prior NTIA sidestepped this by simply saying satellite doesn't count as broadband.

So it's not, we don't have to worry about the eligibility. The current NTIA is saying satellite does count as broadband, but it doesn't count as broadband in terms of servicing a place already, and therefore you can overbuild yourself with this broadband service. That is super awkward to begin with.

Second, have, and Tennessee is a great example of this map, so you have this heterogeneity, this skewing of where the proposals are. So you've got like 25 % of the state that's got no proposals. You've got 25 % of the state that's got like only LEOs, satellite provider proposals. And then you've got this small group that's got like a whole bunch of providers that are all in the mix together. Probably places where fiber networks are expanding their existing footprint.

Here's the problem.

you're not gonna bridge the digital divide. If you look at this map, you're like, well, Tennessee is still screwed. This didn't solve their fundamental problem to begin with in terms of having a universal broadband solution. But in all the areas, pretty much, and I haven't checked every single county, so maybe this is not correct, but I believe in just about all the areas, LEO is way cheaper than any other. So like,

Ry Marcattilio (30:16)
I saw

two counties where ⁓ a fiber provider was under bidding LEO and it was the municipal network in that area and I assume it's because they're you know right next door and we're talking. Yeah.

Sascha Meinrath (30:25)
line extension. Yep.

That's right.

Christopher Mitchell (30:28)
But overwhelmingly, you

saw the same thing, which is that LEO is coming in way under budget.

Ry Marcattilio (30:33)
Way under 10x less.

Sascha Meinrath (30:33)
that's

yeah and we're assuming that they they're serving all the BSLs in those all the eligible households in those areas so like if it is at apples to apples comparison that means that the state will have to do

unless they step in and say, our analysis show that you can't provide broadband connectivity to all these eligible households because of the density problem that we have documented using math and equations. if we have it wrong, again, we're happy to correct it. math, math is math. So if it's not wrong, then this is a fundamental problem.

Sean Gonsalves (31:05)
Right. Now, for me and for our listeners who may not know, mean, so, BEAD grants obviously are to pay for building infrastructure. If LEO gets a BEAD grant, what exactly are they getting the money to do?

Sascha Meinrath (31:19)
You know, that is a more complicated question than one might think because they're obviously not launching a whole bunch of new satellites. There's a whole... Well, there's a... But this isn't paying for that.

Christopher Mitchell (31:25)
But they are. mean, that's the goal, so, I mean,

right. Well, they're saying it's reserving capacity and money being fungible. feel like it probably is. basically, you know, if I think the answer to Sean's question, from the point of view of people that are bending over backwards to try to figure out how to make this work within the existing law.

Sascha Meinrath (31:33)
That's right.

Christopher Mitchell (31:47)
is that that money will go to increasing capacity and that there will be some kind of SLA that I think you've referenced as well, Sascha, in terms of requiring them to provide a level of service that meets the law's definition.

Sascha Meinrath (31:59)
Yeah, I I've read the appendix that's tied to the notice and can't make heads or tails as what these tax dollars are paid for. Like, reserving capacity on a pre-existing network, like, I don't know what that is. Like, if I have a fiber network, am I just like, ⁓ this strand I'm going to reserve for, like...

Christopher Mitchell (32:16)
Right. Well, I I think the generous

reading of it is going from 8,000 to 12,000 or 16,000 satellites so that they can provide a higher level of capacity. mean, this gets into the question that I wanted to come back to, Sascha, which is that, you you called it a Ponzi scheme. And I'm not going to argue with that, but I will say that, like, if you could get in kind of early and I don't know if we're at the end, you know, that you want to go to like 40,000 satellites or something like that.

Sascha Meinrath (32:26)
you

Christopher Mitchell (32:41)
So maybe ⁓ this will work out. Now, I want to ask some questions about that, but I want to first get your reaction to the idea that, yeah, it's a relatively small constellation relative to what he plans. Will that not solve much of the issue?

Sean Gonsalves (32:54)
So you're telling me that the BEAD grant money will pay for rocket fuel.

Christopher Mitchell (32:57)
Money being fungible, I would expect that yes, it would. mean, right now, think Elon Musk is good at raising money. That's the primary thing that he's good at doing, right? And he raised like $10 billion for SpaceX. And then he raised more money from investors for Starlink. And so that money has been paying for the rocket fuel. Presumably, as they run out of that money, which in my understanding is they you know, they're closer to the end of that money than they are at the beginning.

⁓ then they would be using these funds, I would imagine, to be paying for their costs to deploy satellites and that sort of thing. Sascha, I don't know if you have a different sense of that.

Sascha Meinrath (33:30)
Yeah, again, like I've read through, they are nebulous as to what you're actually buying with tax dollars when you're paying it into that pot. So it's not clear to me what we are purchasing. It's certainly not clear to me that we're purchasing like new build, right? It seems to me that, no, we don't have to be, they can...

Christopher Mitchell (33:44)
Well, I think we have to be. so I mean, to some extent, we know that, well, we know that like a third

well over the next several years, several of the satellites that are up there are going to be ⁓ turning into little fireballs and they will need to be replaced. And so presumably some of the money is going to that. mean, Starlink is bringing in, I believe the revenues are in the billions of dollars now, which as we understand it is not yet self-sufficient, although we don't have any sense of the costs. As you said before, like everything is speculation when it comes to this.

of what the real costs are of launching this stuff.

Sascha Meinrath (34:12)
Well, except again there's

this statutory mandate. You have to buy infrastructure that's being built out to the homes of the BSLs, of the underserved households across the country that the money is for. That breaks down completely in this satellite giveaway. And it's not clear to me that you would actually be purchasing a satellite launch or anything. Like this reserved capacity is this nebulous thing to say like,

capacity

on our pre-existing infrastructure. We're not actually building new infrastructure. It's our pre-existing infrastructure. I'm like, so you're over building yourself. Like you already reached that household and I'm now reserving capacity on your pre-existing infrastructure. Like that's not legal. Right? And you can...

Sean Gonsalves (34:57)
Would some

of the money I would have, is some of the money going to purchase? I mean, it's probably a fraction, but it's some of the money going to purchase the equipment, the install equipment for subscribers.

Christopher Mitchell (35:06)
I would presume

so. I've seen some stories suggesting that the homes that qualify would get a terminal at no charge in some states. I'm not sure if that's the case in all the states. I assume they're striking their own bargains and some states will prove to be better than others. Sascha?

Sascha Meinrath (35:19)
Yeah, no, again, totally nebulous. Like, what are we purchasing? I don't know. I'm an expert in this space and I can tell you, I don't know what the tax dollars are going to. It seems to me that you are, by very definition, overbuilding yourself. That you are reserving capacity on your existing. And you read this appendix to the funding notice and it's like, you're reserving capacity on your existing infrastructure for new users. And I'm like, so can AT &T do that?

like hey we're just going to pay for our own pre-existing infrastructure to reach the same households we existed that we already reach. I think people would see that as obviously crazy.

Christopher Mitchell (35:57)
Well, the reason that... Right, no, deeply

corrupt. Now, we've run to great lengths ⁓ to try to avoid doing that.

Sascha Meinrath (36:04)
Not just corrupt, illegal. You can't

overbuild with this funding. It's a statutory requirement that you don't do that. And so how they are...

Christopher Mitchell (36:13)
Which is funny because it's the people

that pushed for that are the ones that want this solution. Now, to some extent, this situation in Tennessee brings further question to a lot of this because how do you read the satellite companies not bidding on a substantial number of BSLs where they would have been the only option? Is that a tacit admission that either they're either they're leaving money on the table, which you wonder why they would do that. Or there's a tacit admission that there's a

Sascha Meinrath (36:17)
That's right. yeah. I mean, hoisted by their own petard. Yes.

Christopher Mitchell (36:41)
ton of homes that they just cannot serve.

Sascha Meinrath (36:43)
Yeah, I'd love to get sort of tree coverage and topological maps of those areas where nobody bid and see if there's a commonality there. Again, this is wild hypothesizing here, but I suspect it's a combination of either pre-existing density, although I doubt that in these areas because of the poverty levels. I think it's a question of tacit acknowledgement that satellite doesn't work well in these particular regions of the country, i.e. hilly regions and tree

covered regions and et cetera. Yeah, so it's like actually I guess three variables here. Somebody can run the analysis. It's gonna be trees, topology, and poverty.

Christopher Mitchell (37:13)
Someone should tell West Virginia that.

I want to talk about the thing that drives me nuts that nobody else talks about, feel like, which is the assumption that we can just put another or that it will be affordable and reasonable for SpaceX to launch so many satellites. I think that there are significant issues that may come up, whether that is cost. think SpaceX's assumptions about future costs are deeply problematic as Starship cannot get off the launch pad in return.

⁓ I think there's reason to believe that they're not going to solve that anytime soon. In the reporting in New York Magazine, I believe they suggested that the V3 Starlink satellites can only be launched by the Starship. They cannot launch on Falcon 9s, which is another challenge. So, Sascha, I'm curious if you've been digging into any of that.

Sascha Meinrath (38:07)
I have and actually if you look at the de-orbiting numbers more and more of them are coming down as well But you know again what what you're describing Chris is a Ponzi scheme right that the cost structure only works as long as you keep this sort of like topsy-turvy addition of new users etc, but the burn rate is tremendous and again We haven't seen the numbers to explain how this is sustainable over the long term

It's not clear to me that it is, and again, this is at the edge of my ken, so I want to be cautious in overstating my skepticism. It's certainly not sustainable as it is currently architected for the provisioning of broadband at scale, which if that was the goal, is to pay for this thing via a large user base, I think we're going to find that that's not going to work on a variety of levels.

into that mix, I would give you Starshield. Which is to say, once you've stopped sucking in on the teat of public funding for residential services, there's always the military industrial complex. And that seems to me the place where were I a Machiavellian Ponzi scheme schemer. That's where I would look next.

Christopher Mitchell (39:24)
There has been writing suggesting that ⁓ some of the NASA folks that have made significant decisions to steer large contracts to Starlink ended up working at Starlink with generous packages, let us say. I'm not sure how long that continues. I think other space companies may figure out the corruption route as well. So I have deep concerns about the ability to launch all that stuff, but I will say that like...

It doesn't strike me as impossible. And I don't think you're saying it, but I want to make sure that people understand this because I think you would leave open the possibility that Starlink continuing to draw in revenue around the globe could generate enough funds to be a sustainable from a financial perspective, if not environmental, because we don't even know what the impact on the environment is of all of these launches and degradations. We're just starting to see the science on that. But it may be financially sustainable. We don't know that that is true.

at all yet. A lot of people seem to just take it as a given.

Sascha Meinrath (40:20)
Yeah, I mean again, there's some complexity here in terms of developing the sustainability math behind this because it's both the maintenance costs the operating expenditures for something brand new So we don't really know like and this this involves everything from like what's the Sun flare cycle? Because once the earth's atmosphere expands then things deorbited a higher rate like this There's a lot of unknowns here that we will only learn about kind of as they happen

But I would certainly agree, and I think it's really important to point out that Starlink can provide really incredible services that maybe should be subsidized. Which is to say, like, the function of having an always-on satellite communication system on a global scale is, that's worth something. But that might...

Sean Gonsalves (41:07)
Well,

the price to a subscriber, especially in low income areas could certainly be subsidized. I Starlink is not, service is not cheap.

Sascha Meinrath (41:16)
That's correct. mean if we had to pay per mile for the actual cost of roads, nobody'd drive anywhere. So like there's certain mission critical infrastructure that we understand as a public good and we subsidize as such. We're not even close to that in the telecommunications.

And I would be very skeptical of doing that, say, on the basis of the corrupt things that are happening and the ways in which that we are privileging funneling public funding to Starlink at this point in

Christopher Mitchell (41:43)
Ry, we're running out of time. Which one was last question?

Ry Marcattilio (41:45)
It wasn't a question actually, it was just to underscore that point that Sean just introduced, which is that Starlink service is not inexpensive. It's 120 bucks a month for standard, it's 80 bucks a month or something for the light. And I just wanted to point out that.

Christopher Mitchell (41:57)
They did introduce

the low cost option in New York state. So I feel like this is also an ongoing conversation. I just want to throw that in, Ry.

Ry Marcattilio (42:04)
Yeah, that's fair enough. I was going to make the point that at the end of the day, maybe everything works out OK, or mostly the way that we might hope, given how bad it looks now. But then you've got millions of new households on the Starlink service, and they're paying more for less service than everybody else. If you look at your average wireline provider in the United States, and I wouldn't ⁓

you know, take these numbers to the bank without double checking them, but you're paying on average like 10 or 20 mega, you're paying 10 or 20 cents per megabit per second per month for that service. So you're paying like 80 bucks a month for gigabit service or, you know, or 110 for two gigabits or whatever. In comparison, the Starlink service is like 80 cents to a dollar per megabit per second per month. So it's four times to whatever, eight times as much to connect, which

You know, some families can afford to pay, but a lot, a lot, a lot of families cannot.

Sascha Meinrath (42:57)
What?

Christopher Mitchell (42:57)
And that doesn't even touch on the latency issue that

Sascha raised. Sascha raised the issue of the six terminals per, six to seven terminals per square mile. I wanted to note before we got out, Sascha, I'll give you the last comment in a second, but we've seen this empirically as well. If you look at my understanding is first responders discussing when they were at the Rose Bowl and responding to the LA fires, a lot of them brought their terminals and they barely worked.

If you look at the response in Texas, first responders brought their low Earth orbit, know, Starlink terminals, and they were getting one to two megabits per second, according to one of our loyal fans that monitors the sites and keeps track of those things, sent me a note. I I didn't do that research myself, but there's empirical evidence out there that when these terminals get together, that they definitely are not able to deliver, and we're not even talking about like a service that's like a little bit slower.

basically unusable for the function that it's there for at that point at one or two megabits a second. Sascha?

Sascha Meinrath (43:55)
Yes, so two things, you know, I spent a number of years as a first responder in a prior life and you we did things like responding to Hurricane Katrina, Superstorm Sandy, the hurricane or the earthquake in Haiti. Consistently infrastructure that was five-nines reliability fell down.

So when I'm like skeptical, because like I've lived through multiple iterations of this problem, but I want to maybe doubly underscore what Ry and Sean bringing up, which is like, if we're going to subsidize telecommunications, which I think we should, we do healthcare, we should do education, we should do telecommunications. I think that's what living in a 21st century civil society looks like. To avoid waste, fraud and abuse, you have to do the all in longitudinal analysis of the total cost.

of this infrastructure. And the fact remains, yes, like fiber's gonna be more expensive on the front end, but the cost, the operating expense for fiber is like tiny. So once you make that initial cost, it's cheap. And so if we're subsidizing this infrastructure, what I want is the all-in best option. And it turns out, it's not just speeds and reliability that fiber's better at. According to all analyses that I have seen thus far,

the lifetime, let's say 30 years of infrastructure, fiber beats the crap, kicks satellite to the curb. Which isn't to say the niche of satellite doesn't exist, it does. It's rather to say it is a foolhardy public decision to subsidize the worse, more expensive option, vis-a-vis cheaper and better service provisioning.

Christopher Mitchell (45:35)
Yeah, I will just, I'll close this out by saying it drives me crazy to see the people who are supporting the way the Trump administration has gutted this program for rural America to turn it into a program for low-earth orbit satellite, because they continue to say it's just not, it's not affordable to deliver fiber to all these areas. When we have more than half of the states, we're given, all of the states were given a budget and more than half of them, we know they came back and said we can reach 75%, 80%, 90%.

of locations with fiber within the budget giving everyone service. These are things that I didn't think was possible a year or two ago. I thought we were gonna get much worse numbers on these plans. And in fact, we gave states a budget. They came back and did it.

And now a bunch of people just want to pretend that didn't happen and say, well, we couldn't possibly do fiber to 90 % of the locations. It would never work. Despite the fact that we have the maps, the engineering, the freaking contractors were lined up to do it. And we're acting like it's an impossibility. Not everyone, most people understand this, but like those people that are commenting on LinkedIn are driving me nuts.

Sascha Meinrath (46:36)
Yeah, it's a boondoggle of our own making and that has a rich history in federal funding of broadband.

Christopher Mitchell (46:38)
Sascha?

Yes. Yeah. I feel like we almost had the one program that we could be proud of. It was a little bit of a mess. It was a lot of a mess to get started. And we probably, you and I, you and I've talked Sascha about how we could have done it better, but frankly, it would have been one of the best programs. And finally, we could have stopped talking about rural infrastructure and telecom for the most part, but instead we have just gone for the own goal. It's a total own, you know, a total own goal. I was trying to think of the tennis term, the

unforced error. ⁓ Sascha, thank you so much. I hope that the rest of your Canadian vacation is slash summer is wonderful. I don't know, maybe we'll find an excuse in two weeks to bring it back.

Sascha Meinrath (47:14)
you

It's great up here, but the broadband sucks. It ⁓ turns out the lack

of competition up here has the same impact on quality of service as it does in the United States.

Christopher Mitchell (47:34)
Yeah, no, and

I was just talking to someone about the Alberta Supernet, which is a bit west of where you are, where there was a government program. And that was also riddled with kind of like a sweetheart deal and corruption and ended up being not nearly the benefit to the public that it was supposed to be. So we are not alone in the United States, at least. Thank you all. Thank you, Ry. Thank you, Sean. Did you want to say something, Sean?

Sascha Meinrath (47:54)
you

Sean Gonsalves (47:56)
Captain's

Log, star date, 7-24-25.

Christopher Mitchell (47:59)
And Sascha, thank you one more time.

Sascha Meinrath (47:59)
Very nice.

Ry Marcattilio (48:06):
We have transcripts for this and other podcasts available @communitynetworks.org/broadbandbits. Email us at podcast@communitynetworks.org with your ideas for the show. Follow Chris on BlueSky: his handles @SportsShotChris. [00:39:00] Follow Community Networks stories on BlueSky: the handles @communitynets. Subscribe to this and other podcasts from ILSR, including Building Local Power, Local Energy Rules, and the Composting for Community Podcast. You can access them anywhere you get your podcasts. Catch the latest important research from all of our initiatives if you subscribe to our monthly newsletter @ILSR.org. While you're there, please take a moment to donate your support in any amount. Keeps us going. Thank you to Arnie Hesby for the song Warm Duck Shuffle, licensed through Creative Commons.