Hacker News Reader

Your GPS receiver is not going to be able to "see" the current satellites. All of the other people who share this spectrum do so "horizontally" if you will, this frequency band isn't reflected by atmospheric effects so omni-directional emitters are not visible to antenna that are looking "up" to the sky. The problem with this proposal is that the company wants to broadcast DOWN from satellites which these antennas will see and it will reduce their sensitivity to neighboring signals in the same band. Since you need 4 - 6 satellites for a good tracking lock on GPS, the likelyhood is that you won't be able to get that because of inter-band interference.

Can you fix it? Sure, you can add a $10 - $50 front end selector on the GPS devices to increase their selectivity to GPS only bands. That will raise the cost of "having" GPS across the board and legacy devices won't see the benefit so they will become useless.

It is hard to imagine that any technical voices at the FCC thought this was a good, or even reasonable, idea.

And on a humorous note, "U.S. Space Force Gen. John Raymond." I chuckle that the Space Force already has generals but of course it does, but not ones that "came up through the ranks, from Spaceman 1st class."

Chuck your comments are always filled with great information and humour. Thank you for your years of contribution to HN so that lurkers like myself can be informed and entertained.

I'm not an RF person, but the way it was explained to me is that GNSS signals are in the neighborhood of 1560mhz-1610mhz. With the advent of GPSIII, even more of the important signals are getting pushed outwards towards the extremes of that range. Ligado/Lightsquared initially wanted permission to basically surround GPS (1526-1559mhz and 1610-1660mHz)

After lots of complaints, they amended their request to leave a buffer of 20mhz or so between their signals and GPS. However, Ligado wants to transmit at ~10watts, and the GPS signal is in the neighborhood of a femtowatt by the time it hits a receiver. That can apparently still cause interference. For the GPS chip in your phone it may not make a huge difference. However testing has shown degraded performance in aviation GPS receivers more than a mile away from a 10watt transmitter. For even higher precision receivers (think surveyor equipment or maybe an autonomous tractor) the radius of the degradation stretches more than 2 miles from the transmitter. Ligado wants to put their transmitters ¼ of a mile apart.

Source for some of those numbers here, and an entertaining talk if you're ever able to see it in person: https://rntfnd.org/wp-content/uploads/Brad-at-UAG-Users-Advi...

Why can't we just greatly improve the signal strength at source by an order of magnitude? Like I know the new satellites can use beam forming to increase power for some large area for military use but it looks like even so total power is about ~500W right now. What's stopping us from deploying 10KW sats?

>Why can't we just greatly improve the signal strength at source by an order of magnitude? Like I know the new satellites can use beam forming to increase power for some large area for military use but it looks like even so total power is about ~500W right now. What's stopping us from deploying 10KW sats?

Even assuming that the very smart and capable people who developed GPS haven't thought about that already, there is zero-point-zero percent chance that the government will (or should) turn 24 satellites that cost 500 million dollars each into very expensive flying paperweights, for the benefit of some tiny 5G startup.

Deploying a 10KW sat is not a problem. Replacing a whole fleet just to keep it operating the same way as before is.

Supplying the power to the satellite, and getting the heat out. The RF amps are at most 50 % efficient, so you are dissipating 1/2 the input power as heat.

> Like I know the new satellites can use beam forming to increase power for some large area for military use

No, what you think you 'know' is plain wrong. Usually when discussing satellites, we're talking about a directional antenna that focuses a spot or wide beam for broadcast. A spot beam allows you to target a particular region, DirecTV uses spot beams to provide local broadcasts to general regions of the US. The military uses M-code shit on GPS with a directional antenna for the same sort of use case.

Beam forming is a different concept where you get creative with multiple antennas broadcasting at once and exploit signal cancellation to improve the data rate by shifting the phase between them. It does increase your received signal but relies on the receiver transmitting information back to the satellite (which is not going to happen for GPS, at least not the vast majority of consumer equip). Beam forming discussed on HN just last month: https://news.ycombinator.com/item?id=22784132

> What's stopping us from deploying 10KW sats?

Sat power is around and above that already, it doesn't make sense to increase the spot transponder ERP too much or you end up with issues like interference with adjacent bands since RF gets trickier to filter the higher the power it gets. Keep in mind a comm satellite usually has lots of transponders using maybe 100-500W each, so it adds up to some geosynchronous ones being like 20kW total consumption. Smaller LEO ones obviously not going to be close to that.

GPS released a paper about interfering with itself: https://www.gps.gov/multimedia/presentations/2010/10/ICG/bet...

Communications history is littered with failure stories of people who wanted to throw power at the problem, when they only needed to find ways to listen better.

Thanks that's actually helpful. I did some more reading and it looks like it is as you say, the new military transmitters are directional and steerable, beam forming done on the receiving side.

But still it seems that power is still part of the solution given the increased power density of the new military antennas. I guess what I'm wondering is why they can't just reserve more band, move to regional antennas for whatever other technologies in addition to fighting polluters/jammers/spoofers and noisy neighbors. Or is that the plan?

> the GPS signal is in the neighborhood of a femtowatt by the time it hits a receiver.

This sounds more dramatic than it is - my WiFi signal from less than 10 feet away is approximately 1 nanowatt, or approximately 10^8 times weaker than when it left the AP radio. Thats just how RF works.

It's still pretty dramatic when you compare them directly. A 500watt transmitter that's 30,000km away is going to be many orders of magnitude (11?) weaker than a 10watt transmitter that's only 1km away.

When light squared was shut down for messing with gps frequencies in their satellite network, the military did not say that light^2 was actually broadcasting outside of the expected frequency band. The deal is that the military just doesn’t use filters on most of their broadcast and reception equipment so their operating frequency is actually way wider than they say publicly.

This is because filters cost money and reduce the effectiveness of the broadcast system. If you could get away with operating over a much wider spectrum for any service, you probably would.

This is why when the military says “this interferes with gps even though it’s outside of the gps band were supposed to be using” what they mean is that they are using a much wider band of spectrum than what’s in the specs.

Ultimately, the fcc doesn’t have regulatory authority over the dod.

> The deal is that the military just doesn’t use filters

Do you have a citation for that?

I would have thought the issue is not that GPS receivers don't use filters, but rather than no filter has a perfectly sharp cut-off and GPS signals are exceptionally faint.

For example, if the GPS signal is 0.1 femtowatts at 1575MHz, even if you've got a filter that can remove 99.99999% of a noise signal at 1600MHz you can still overwhelm the GPS signal with 1 nanowatt of power.

Oh man, I bet there's a nice market of GPS jammers out there - especially useful for the home team in any action, since they don't need help with directions. I wonder if that counts as a weapon?

Some truckers use them for manipulating the trackers on their vehicles. Some of them are stupid enough to have them active when they are anywhere near an airport, which makes it far more likely they'll get caught.

Precisely -- no RF filter has a perfect "brick wall" response, and no transmitter will have a perfectly pure output either. GPS receivers already have to use some pretty crazy tricks to pull a signal out of the noise; it doesn't take a lot of interference to make that impossible.

There is also the insertion loss and physical size of the filter to consider. It's a four variable problem where you can only optimize for any two.

That's not accurate.

Every GPS receiver in existence uses filtering. The question is, how sharp does that filter have to be? The GPS frequencies were chosen to be away from frequencies used by transmitters that might be near GPS receivers. Lightsquared proposed repurposing satellite downlink frequencies (which they bought at a discount because this restriction limited their use) for the purposes of terrestrial networking, breaking that assumption. In order to maintain similar functionality, GPS receivers would require significantly sharper filters, increasing bulk, cost, and complexity.

Thanks, this comment clearly explained why there even is an issue

Thank you so much for sharing this. I had heard the story from one of the Lightsquared guys at a conference but didn't get the details.

I appreciate this comment so much!!

>everyone knows that would break GPS (which already operate well below the noise floor).

Can you explain the noise floor bit?

There exists a fundamental level of electrical noise at all frequencies typically controlled by the background temperature that sets the minimum level of signal you can resolve before it gets washed out in noise. GPS uses multiple coherent (noise being random instead of coherent does not combine at the same rate) copies of the signal that can be recombined to increase their effective signal, so even though they are transmitted below the universal noise floor, they are above it once recombined.

*edited for spelling

Thanks, that's very interesting. Appreciated.

The reply that you already have is pretty good. https://en.wikipedia.org/wiki/Gold_code is a good jumping off point for further research (as is the wikipedia article on GPS).

Perhaps it's not about the nominal operation as a 5G tower; rather, a bunch of powerful software-defined radios that could interfere upon command is a pretty reasonable attack vector.

> Did the FCC stop Amazon from importing these things?

Did you report it to the FCC so they'd know?

Ligado’s biggest investors are JPMorgan, Centerbridge, and Fortress. David Redl, the former head of the NTIA, supported Ligado’s mission while the GPS (part of DoD but also co-run by the DOT) was against it for as long as this issue existed. The FCC pushed this through because Ajit Pai is retiring at the end of the year (regardless if Trump wins) and wants a legacy of approving more commercial uses of spectrum.