An amateur satellite tracker stumbled across the signal, which is coming from Starshield satellites in a "hidden" part of the radio spectrum.
A constellation of classified defense satellites built by the commercial company SpaceX is emitting a mysterious signal that may violate international standards, NPR has learned.
Satellites associated with the Starshield satellite network appear to be transmitting to the Earth's surface on frequencies normally used for doing the exact opposite: sending commands from Earth to satellites in space. The use of those frequencies to "downlink" data runs counter to standards set by the International Telecommunication Union, a United Nations agency that seeks to coordinate the use of radio spectrum globally.
Starshield's unusual transmissions have the potential to interfere with other scientific and commercial satellites, warns Scott Tilley, an amateur satellite tracker in Canada who first spotted the signals.
"Nearby satellites could receive radio-frequency interference and could perhaps not respond properly to commands — or ignore commands — from Earth," he told NPR.
Outside experts agree there's the potential for radio interference. "I think it is definitely happening," said Kevin Gifford, a computer science professor at the University of Colorado, Boulder who specializes in radio interference from spacecraft. But he said the issue of whether the interference is truly disruptive remains unresolved.
SpaceX and the U.S. National Reconnaissance Office, which operates the satellites for the government, did not respond to NPR's request for comment.
Caught by the wrong antenna
The discovery of the signal happened purely by chance.
Tilley regularly monitors satellites from his home in British Columbia as a hobby. He was working on another project when he accidentally triggered a scan of radio frequencies that are normally quiet.
"It was just a clumsy move at the keyboard," he said. "I was resetting some stuff and then all of a sudden I'm looking at the wrong antenna, the wrong band."
The band of the radio spectrum he found himself looking at, between 2025-2110 MHz, is reserved for "uplinking" data to orbiting satellites. That means there shouldn't be any signals coming from space in that range.
But Tilley's experienced eye noticed there appeared to be a signal coming down from the sky. It was in a part of the band "that should have nothing there," he said. "I got a hold of my mouse and hit the record button and let it record for a few minutes."
Tilley then took the data and compared it to a catalog of observations made by other amateur satellite trackers. These amateurs, located around the world, use telescopes to track satellites as they move across the sky and then share their positions in a database.
"Bang, up came an unusual identification that I wasn't expecting at all," he said. "Starshield."
Starshield is a classified version of SpaceX's Starlink satellites, which provide internet service around the world. The U.S. has reportedly paid more than $1.8 billion so far for the network, though little is known about it. According to SpaceX, Starshield conducts both Earth observation and communications missions.
Since May of 2024, the National Reconnaissance Office has conducted 11 launches of Starshield satellites in what it describes as its "proliferated system."
"The NRO's proliferated system will increase timeliness of access, diversify communications pathways, and enhance resilience," the agency says of the system. "With hundreds of small satellites on orbit, data will be delivered in minutes or even seconds."
Tilley says he's detected signals from 170 of the Starshield satellites so far. All appear in the 2025-2110 MHz range, though the precise frequencies of the signals move around.
Signal's purpose in question
It's unclear what the satellite constellation is up to. Starlink, SpaceX's public satellite internet network, operates at much higher frequencies to enable the transmission of broadband data. Starshield, by contrast, is using a much lower frequency range that probably only allows for the transmission of data at rates closer to 3G cellular, Tilley says.
Tilley says he believes the decision to downlink in a band typically reserved for uplinking data could also be designed to hide Starshield's operations. The frequent shift in specific frequencies used could prevent outsiders from finding the signal.
Gifford says another possibility is that SpaceX was just taking advantage of a quiet part of the radio spectrum. Uplink transmissions from Earth to satellites are usually rare and brief, so these frequencies probably remain dark most of the time.
"SpaceX is smart and savvy," he says. It's possible they decided to just "do it and ask forgiveness later."
He notes it's unlikely the signals from Starshield have caused significant disruptions so far, otherwise other satellite operators would have complained.
Tilley told NPR he has decided to go public with his discovery because the world's satellite operators should be aware of what's happening.
"These are objects in classified orbits, which could potentially disturb other legitimate uses of space," he said.
Read the original article
Comments
By anonymousiam 2025-10-1715:551 reply The S-band uplinks are not typically used during operations. They're mostly used during transfer orbit operations and initial testing, and in emergencies when something goes wrong with the normal comms (safe mode). The S-band antennas on the satellite are typically omnidirectional, so they'll hear anything strong enough to overcome the noise floor. Those comms can be encrypted or in the clear, depending upon the situation. The military satellites that I'm familiar with stop listening to the S-band uplink when their normal uplink is operational, so interference shouldn't be an issue during normal operations.
I'm not involved in this stuff anymore (now retired), but it's possible that the Starshield constellation supports transmitting on S-band (or L-Band) as a means to relay SGLS communications to satellites that are out-of-view. Having this capability would greatly benefit the workflow of transfer orbit operations and initial testing, by eliminating the constraint that the satellite must be in-view to communicate with it. It would also benefit anomaly resolution by allowing instant access to a malfunctioning spacecraft.
https://en.wikipedia.org/wiki/Satellite_Control_Network
https://www.orbitalfocus.uk/Frequencies/FrequenciesSGLS.php
https://www.af.mil/News/Article-Display/Article/473264/af-sp...
By stinkbeetle 2025-10-1717:171 reply > I'm not involved in this stuff anymore (now retired), but it's possible that the Starshield constellation supports transmitting on S-band (or L-Band) as a means to relay SGLS communications to satellites that are out-of-view. Having this capability would greatly benefit the workflow of transfer orbit operations and initial testing, by eliminating the constraint that the satellite must be in-view to communicate with it. It would also benefit anomaly resolution by allowing instant access to a malfunctioning spacecraft.
That's interesting, thank you for the great comment. Would that kind of usage then not be counter to the standards, as suggested in the article?
By anonymousiam 2025-10-1717:25 Any sort of innovation may counter "standards", but without knowing the specific ITU terms for S-band, I can't say whether or not any regulation has been violated.
The fact that somebody saw something pointed at Earth on a frequency generally reserved for uplinks doesn't necessarily mean that it would interfere with other spacecraft receiving the signals from the ground. Starlink (and presumably Starshield) operates in LEO, below most other LEO spacecraft. Maybe they're using a dish or even a phased array antenna, and pointing down instead of up. If so, the probability of interference is low.
The mass of the starlink satellites has doubled twice since V1.
It is not hard to assume that there is a significant DoD rideshare payload involved on the existing commercial satellites. Having a sensor platform on every single one would be incredible. The satellites that have been officially branded as Starshield (~183 we know of) could be part of cover or a more "kinetic" mission profile.
If I was in charge at the Pentagon, I would want every one of those 10k birds to have my sensor package on it. I also don't think I would permit a commercial spaceflight vendor to perform as many launches as SpaceX has performed without some kind of arrangement like this in place.
By QuiEgo 2025-10-1717:28 > would permit a commercial spaceflight vendor to perform as many launches as SpaceX has performed without some kind of arrangement like this in place.
I’m sure SpaceX is happy to take the DoD’s money, doubt there’s any strong arming needed.
It's an interesting theory, but I'm not sure it works. Do you design in these sensors and build in the normal starlink factory, where most people don't have security clearances? Or do you make thousands of people get security clearances, making it pretty obvious?
And as soon as any data from a specific sensor leaks, adversaries would likely be able to pinpoint what satellite produced it.
And then the contractual terms mandating commercial spaceflight vendors do this work.
It all gets really complicated, with many thousands of people who are not part of traditional intelligence services all having to keep a massive secret.
> where most people don't have security clearances
Are we sure about that? Because i see plenty of clearance required job listings to their redmond facility.
> And as soon as any data from a specific sensor leaks, adversaries would likely be able to pinpoint what satellite produced it.
Yeah. Thats why usually data from any satelite would be very closely held. Even with old satelites it was a big deal when the president just posted an image publicly.
But i don’t understand your argument. Even if it leaks that wouldn’t make the sensor network worthless. Like this argument is true for any spy satelite. If the data from any sensor leaks it is bad. Not a reason to not make the satelite.
You’re saying that thousands of starlink satellites are actualky spy satellites, but it’s scret because spacex hires some people with security clearances, but it doesn’t matter because it doesn’t have to be secret to be useful, but it is secret despite involving 100x times more volume than any previous spy satellites, and the only evidence is… vibes?
By krisoft 2025-10-189:37 > You’re saying that thousands of starlink satellites are actualky spy satellites
No. I didn’t say that and wouldn’t say that. Read carefully because i write carefully what i mean.
You are saying that the theory as presented by bob1029 could not work. I’m saying that your argument why it couldn’t work is not persuasive.
You present two arguments in your comment. (As best as i can understand it.)
One is about the secrecy around the design and manufacturing of the satelites. You claim, without support, that most people don’t have a security clearance in the redmond starlink factory.
Satelite design and manufacturing is already very secretive. Because of ITAR and regular commercial confidentiality you won’t hear a peep about what is on the satelites. The people who design and develop the satelites would of course know the full capabilities of them, but the people manufacturing them need not know. All they need to know is that they are installing optical assemblies. Whoever asks can be told that they are for laser communication.
So the amount of people who need to know is smaller than the full work force. That workforce is already trained on secrecy, and they are practicing it. They already risk prison if they leak anything. (Without any spy satelite business, just because what they work on is ITAR controlled.) On top of that spacex is quite openly hiring for a number of positions requiring top secret clearance.
Your other argument is that if data from these hypothetical sensors would leak that would compromise the hypothetical secrecy around them. Which is true, but is a general property of all inteligence gathering. If it was not an argument against any of the other systems why would it be an argument against this one?
Secrets like this hypothetical one have a finite lifetime. You do it because you hope to gain from doing it. You keep it secret because you hope to gain more than if you didn’t keep it secret. Even if the capability becomes known to your adversaries you won’t loose all the benefits, just some.
> and the only evidence is… vibes?
Because i’m not claiming that they are spy sattelites. All i’m claiming is that your argument claiming that they are not, or couldn’t be is not persuasive. The negation of the statement “they couldn’t be spy satelites” is not “they are spy satelites”, but “they could be spy satelites”.
I hope that helps clarifying what i wrote. Happy to answer any further questions.
By stinkbeetle 2025-10-1716:58 If you were in charge at the Pentagon, would you have the authority to prohibit civilian access to space on that basis?
This reeks of SAR (Synthetic Aperture Radar), and we know Starshield sats carry custom sensor payloads that normal Starlinks don't have.
I've been a pretty harsh critic of Starlink. I don't think it's going to compete well against terrestrial wireless links, specifically 5G FWA. But if they can actually do something interesting with distributed SAR in a large constellation, that's one of those national security breakthroughs that's worth every penny.
The issue with terrestrial wireless links is that you just aren't going to cover every inch of Earth's surface with terrestrial wireless links.
At the same time: coverage comes cheap to Starlink. Which makes it perfect for serving areas no one wants to serve. Such as rural areas, anything outside the largest cities in underdeveloped countries, the open ocean, and so it goes.
Places only Starllink can reach are an insufficient and shrinking TAM. The only places a terrestrial wireless provider doesn't want to serve are places that can't afford FWA even though it costs less.
By stinkbeetle 2025-10-1717:402 reply > Places only Starllink can reach are an insufficient [TAM]
Insufficient for what?
> and shrinking TAM.
Starlink has made quite an impact on planning around servicing commercially non-viable or marginal customers in government and telcos where I am from. It is IMO quite likely that some existing cell towers in remote areas that are very expensive to operate and maintain will eventually be shut down. So that could actually expand the "TAM".
> The only places a terrestrial wireless provider doesn't want to serve are places that can't afford FWA even though it costs less.
No, they also don't want to serve places where it costs more.
By Zigurd 2025-10-1813:16 Satellites that last 5 years and have un-transparent launch economics are very unlikely to win on cost in any land-based market. If the market is too small or too poor to support FWA, it's not going to contribute significant revenue to Starlink.
Starlink needs tens of millions of subscribers to be valued like a telco.
The US military wants to have option to deploy anywhere in the world, so they also want to have comms everywhere.
The starlink network surely has special features to support US military needs (resiliance, encryption, blocking enemy countries from access, robustness against countermeasures, yaddayadda).
By Zigurd 2025-10-1813:17 The military has interoperating and compatible comms everywhere. People have solved many problems before Elon invented a thing.
By ggreer 2025-10-1718:24 5G fixed wireless makes sense in urban and suburban areas because a single tower can serve many customers. But in rural areas, the number of customers per tower is much lower. To be profitable, the cost to build & maintain the tower must be covered by the customers in the area. Google says a 5G tower can cost anywhere from $150k-$300k, with maintenance costs being around 10% of that per year. In rural areas, the cost is likely on the high end because power and fiber lines will be longer. Each 5G tower has a range of 5ish miles, meaning it can serve an area of 75-80 square miles. A rural county like Ferry County, Washington has 2 households per square mile. So in that county, a tower can serve 150-160 households. Assuming 100% of households are customers, that means each must pay $200-400 per year for the tower to break even after a decade. The actual numbers are fuzzier. Obviously you won't have 100% adoption, but cell towers also serve cell phones.
In comparison, each Starlink satellite costs around $1 million to manufacture and launch, and each satellite lasts at least 5 years. So cost per satellite per year is $200k. They currently have 7,600 satellites serving 7 million customers, meaning on average, each satellite serves almost 1,000 customers. At $200k per satellite per year, each customer needs to pay $200 per year for them to break even. It seems likely that launch costs will go down in the future, meaning this number will decrease.
There's also the complication that each new Starlink satellite improves coverage & bandwidth for the entire globe, while each new 5G tower improves coverage & bandwidth in a specific area. A county may have a population density of 2-4 households per square mile, but many of those households are clustered together. The less dense areas are not likely to be covered by cell towers any time soon, as it's less economically viable. Another disadvantage of cell towers is service failures. A single Starlink satellite failure means a slight degradation of service, while a single cell tower failure means everyone in the region is taken offline. In areas where both services are available, people would be likely to prefer the more reliable option.
By robotresearcher 2025-10-1716:061 reply Also where terrestrial access is denied by the local government or infra has been disabled or destroyed.
I hope Starlink can maintain enough regular non-crisis subscribers to subsidize this incredibly helpful use case.
Perhaps governments could/do pay a retainer to keep this option alive.
By jdonaldson 2025-10-1716:56 I'm quite sure that's what's already going on. LEO is big and empty and pretty boring, but it's also resilient to increasing trends of natural disaster, war, and civil unrest. Having a foothold there matters more and more every day.
By oceanplexian 2025-10-1716:51 Once you cross a certain threshold Starlink is superior to fiber optics. That's because it has the potential to be both lower latency and more direct than terrestrial fiber as the speed of light is faster in the vacuum of space.
At a certain scale you're going to have to make the argument that laying a 10,000KM glass fiber across the ocean for 10-20% more latency is a better value than beaming it around in LEO.
By JumpCrisscross 2025-10-1719:59 > Places only Starllink can reach are an insufficient and shrinking TAM
Planes; yachts; cruise ships; naval vessels; sea-based drilling, mining and research platforms; mines in the middle of nowhere are a shrinking TAM?
You may also be underestimating how many large rural landowners don't want to give telcos (and the relevant authorities) access to any of their land.
By fragmede 2025-10-1718:20 The two places that I'm fairly confident will never get covered by terrestrial wireless are the Pacific ocean and Atlantic ocean. Whether yacht owners, cruise ships, and airplanes are enough to sustain the business by some future time when terrestrial customers churn, who knows, but those customers aren't going to go away any time soon.
By pavel_lishin 2025-10-1715:221 reply > Places only Starllink can reach are an insufficient and shrinking TAM.
And yet, people live in those places, and you telling them that they're not economically worth serving isn't really solving their problem.
Pretending that they are worth economically serving with current tech or even tech that could be available soon does not help them either.
By pavel_lishin 2025-10-1715:44 Who's pretending? If they're getting internet through Starlink, it sure sounds like they're being served.
For Starlink, maybe. However, I'm sure nuclear submarines operating in the arctic (and the US Navy in general) as well as forward operating US bases, would love and be willing to pay for Starshield to cover everywhere.
Pretty sure they don't have any comms signals and wouldn't connect to that unless absolutely necessary emergent situation. I thought they operate in the dark.
Subs do pop up to make status reports or check-ins. They do operate independently in that command knows what area they patrolling but not exact locations at any given time.
By anonymousiam 2025-10-180:58 Subs (at least the US ones) also use "strategic" LPI/LPD milsat comms instead of commercial satcom. You don't want your enemy to geolocate your sub fleet whenever they phone home.
By Incipient 2025-10-1715:21 Starlink has plenty of good use cases: - camper van coverage across whole countries/continents - rural farms/communities - global coverage for convenience (eg business travellers) - huge military use
...just depends if it's economically viable
By greenavocado 2025-10-1714:51 Until the next war
By irjustin 2025-10-1716:00 Planes boats etc
By mensetmanusman 2025-10-1718:18 I’m surprised there are technologists that are harsh critics of low earth orbit satellites.
> national security breakthroughs
Can you explain what you mean by this? Still not sure how SAR would fit in here...
By robszumski 2025-10-1714:251 reply If a giant chunk of the constellation can act as a truly huge antenna, what can you get from that? Super high resolution? Seek/dwell time on a target that is effectively infinite?
And we are going to put that in the hands of Elon musk? Are you fucking kidding me?
By krisoft 2025-10-1715:05 Nobody is discussing putting anything in Elon's hand. We are discussing what he already has in his hand, or can grab for himself if he chooses to.
By next_xibalba 2025-10-1714:424 reply Is there a viable alternative?
SpaceX is the only launch provider and satellite operator that is progressing at a rapid pace and driving costs down.
By notahacker 2025-10-1716:451 reply No need for the satellite manufacturer to be the same as the launch provider, and there's nothing at all special about short-lived commodity satellites for LEO constellations. SpaceX is going to be cost-effective at building them given their experience with Starlink, but cost isn't typically a major concern of the US govt, and certainly not a higher priority than concerns about the satellite operator frequently suggesting that access to his satellites might be contigent upon his views on a particular conflict.
> but cost isn't typically a major concern of the US govt
tell that to any project that has had their budget slashed or out right canceled because somebody thought their project was a waste of money. every contractor is bidding unless your name is Halliburton. what's the famous astronaut quote about sitting on top of a rocket built by the lowest bidding contractor?
By notahacker 2025-10-1718:54 > tell that to any project that has had their budget slashed or out right canceled because somebody thought their project was a waste of money.
Their contracts aren't in defense...
Chinese companies seem to be in process of cloning Falcon 9 and even Starlink (Thousand Sails and other constellations).
In the west the Rocketlab Neutron partial RLV and planned Stokes Space full RLV stand out.
And maybe in a few decades even Arianespace will end up with a Falcon 9 class vehicle! ;-)
By JKCalhoun 2025-10-1714:49 > Is there a viable alternative?
Always a good answer. ;-)
2/3 of Falcon 9 launches are for Starlink. No outside revenue. SpaceX continues to require new investment rounds. So the whole "driving costs down" thing might only work until investors expect some actual free cash flow.
There have been 11 test launches of starship. You might've missed the last one because it didn't do anything new, except shedding parts and exploding less. There's a pretty good chance that program will never beat the cost of Falcon Heavy, or that the technology, like multiple refueling flights to get beyond low Earth orbit, is ever made workable.
The last Starship launch was indeed unspectacular because it didn't try pushing the envelope particularly hard. The previous launches were much more precarious, with many fire balls. But I'm a strong believer in iterative development. It's bad PR when everyone can see every failed prototype, but the "design it once, simulate, and make sure the first prototype flies without issues" boxes you in to conservative design decisions.
By ACCount37 2025-10-1716:52 They did push Starship hard enough on reentry that, reportedly, it ended up with multiple holes burned through the metal hull and into the tanks.
It survived that - did that entire "simulated landing" burn and all.
By mrDmrTmrJ 2025-10-1715:58 Well, if 2/3 of SpaceX's current launches are for Starlink (which deploys satellites in LEO), isn't a two-stage, fully reusable vehicle optimized for LEO deployment the thing SpaceX would want to build?
In terms of "free cash flow" expectations, are you aware that approximately 90% of "space" revenue and profit comes from satellite telecom services, with launch services accounting for about 10% of the mix? SpaceX's development of a telecommunications constellation (Starlink) is highly consistent with historical industry patterns of what makes profit in space.
https://brycetech.com/reports/report-documents/global_satell...
By wildzzz 2025-10-1715:54 If SpaceX only had contract money as revenue, they'd be fine but they probably would not be innovating as fast. The investment rounds are to pay for Starlink build-out and Starship.
By marcellus23 2025-10-1716:281 reply You're more worried about that than having it in the hands of the US government?
By epolanski 2025-10-1716:50 How could you even think the opposite to be a better option?
The US does suffer from a serious amount of issues politically (I'm 100% convinced that presidential republics are flawed) but it's still an organization with plenty of checks requiring popular mandate.
No single private individual should ever hold this kind of influence imho, not even if it is Gandhi or a saint and Musk is quite the other end of the spectrum.
By inemesitaffia 2025-10-1820:39 Search "NRO you can't hide"
By JumpCrisscross 2025-10-1719:461 reply > if they can actually do something interesting with distributed SAR in a large constellation, that's one of those national security breakthroughs that's worth every penny
...could comprehensive SAR over the Earth's oceans uncloak submerged submarines when they're under power?
By yencabulator 2025-10-1920:51 Radio doesn't penetrate water well, for practical purposes almost at all.
By ls612 2025-10-1715:02 Starshield with SAR that could support a missile lock would be a completely transformative capability in any pacific war scenario.
The fact we speak about it already sometimes suggests it has been already deployed, sometimes for many decades.
Can it (a SAR sat doing a flyover) somehow be detected from ground?
By ACCount37 2025-10-1713:49 Yeah, I just updated my own comment too for those not familiar with the term.
By cameldrv 2025-10-1717:05 Yeah and the E-7 was recently cancelled. This is what Hegseth said to Congress a few months ago: “The answer is yes. I would. I would file this entire discussion under difficult choices that we have to make. But you know, the E-7, in particular, is sort of late, more expensive and ‘gold plated,’ and so filling the gap, and then shifting to space-based ISR [intelligence, surveillance, and reconnaissance] is a portion of how we think we can do it best, considering all the challenges"
I wonder if Starshield is the platform that is supposed to replace the E-3.
By martinky24 2025-10-1714:361 reply What suggests SAR over downlink...?
By ACCount37 2025-10-1720:32 Weird band for a downlink, and runs continuously and indiscriminately? It being detectable over Canada suggesting the latter. If downlink - what receives it?
