This is really cool! For anyone looking to do a similar project, it's worth noting that HackRFs are pretty low-spec SDRs by current standards: a number of more recent models, such as the XRTX-CS, based on the Lime LMS7002M front-end, achieve far greater tuning bandwidth - you wouldn't have to build a cluster of them, and could get the same throughout for a lower price and complexity!
It is humbling to look at the pictures and read it carefully then don't even understand what the heck he accomplished by doing the bizarre contraption.
The text says the 8 cards are connected to one antenna. To me it sounds like he's doing it because each card only gives him a fraction of the bandwidth he needs to cover.
Edit: I don't understand why he needs that much bandwidth though - maybe he just happens to have the cluster, but I suspect he could get basically a 1/8 slice (full height, but only covering 1/8th of the width of the picture) with only one receiver.
You definitely don’t need 160Mhz bandwidth. Based on other articles on the source blog, it’s something he had and the additional bandwidth let’s you track more spectrum at once.
Outrageously off-topic... but I bet if SpaceX put a camera under the satellite, customers would love getting real-time imagery of their house from the exact satellite listening to their Dishy McD.
I gather you need a license to collect satellite imagery, which is weird. Or to broadcast it, or something. (But what if it is narrowcast, instead?) SpaceX could totally afford such a license.
Speaking of broadcast, the Starlink terminals could be equipped to pick up and cache packets broadcast from a satellite, with a proxy server inside that can deliver cached content. That way, the satellite only needs to send one copy of its Earth View, or the running World Cup, or the new Marvel release, and everybody in the reception cone gets it from the one broadcast, saving bandwidth and system load all around.
(Of course, something in the customer's player would need to know to use the proxy server to get video frames. But that is just software, therefore easy. Ish.)
From a geopolitical standpoint that would cause SpaceX an incredible amount of trouble since they have to get government permission in every country they want to operate in. Lots of countries would not easily grant such permission without major strings attached.
It would be trivial for SpaceX to serve the images only where they are permitted. People in places where they are not could be shown other places they are allowed to see, instead. (Maybe with a caption inviting them to complain to their Dear Leader.)
It isn't about the serving, it is about the ability to take images. Countries are really picky about such things but have trouble controlling in a lot of cases because they have no mechanism to stop it. Allowing RF in / out of there areas of control gives them that mechanism.
You are confused. At present, anybody with money can loft a satellite, and anybody with money can buy pictures from it, with no need to ask anybody's permission.
With Starlink terminals, local governments could make the right to place them depend on agreement not to deliver local pictures to those terminals. Since SpaceX controls both terminals and satellites, that would be trivial to implement. A government might even try to demand that SpaceX not deliver local pictures to anybody else, either. SpaceX might instead choose not to place terminals in such a country, thereby generating local opposition to those pols. But none of this could interfere with delivering other pictures to terminals not so governed, even right across the border.
Anyway, any such pictures would have much lower resolution than the ones already being sold today. So, the only places that would complain and maybe interfere would be those whose pols are equally confused. I do not claim there are no such pols.
> At present, anybody with money can loft a satellite, and anybody with money can buy pictures from it, with no need to ask anybody's permission.
Minor nit: You need to ask permission of the country responsible for your space operations, if it chooses to regulate that. E.g. in the US, NOAA needs to sign off on any US-operated imaging. But generally doesn't care what the photographed countries think of it, although one could in principle imagine restrictions where the US feels it touches national security.
(This was in the press a while back when SpaceX wasn't allowed to publish live feeds from one of their launches, because the right permits weren't in place)
No thank you. SpaceX for some reason some how gets to invade and ruin night skys for everybody and now this to invade privacy.
Ideas like this are so aggrvating. There is non stop privacy invasion and terible data leaks and anti-privacy lobbying by amazon. No need to give more surveillance and public control tools to fickle Elon or the govt.
(I will add for the benefit of certain people that a little camera on the bottom of a satellite will not be invading anybody's privacy. You might be able to just pick out where your house ought to be.)
The author's blog https://olegkutkov.me/ definitely looks interesting as well (I don't see an article specifically about the OP topic yet, but there's a lot of other stuff)
I don't believe that the HackRF cluster is required for receiving the beacons. After the 11 GHz signal is down converted by the LNB, an RTLSDR should be fine for reception. happysat77 on Reddit used a PlutoSDR, but an RTLSDR should be ok, since the down converted signal is around 1600 MHz. The bandwidth of the beacon is only 1 kHz or so.
You don't need a cluster of HackRFs to see a single beacon. But the combination of several HackRFs does give you larger instantaneous bandwidth and the ability to track an individual beacon over a wider range of Doppler shifts and hence for a longer time / part part of it's orbit. And the ability to see a whole bunch of beacons at once.
Wow that's some snazzy dish hardware Oleg's using there. I guess he's planning on more than beacon-tracing. 11 GHz is about a 1" wavelength ... dish size??
