Re: [Starlink] Starlink Reverse Engineering

[Alexandre Petrescu <alexandre.petrescu@gmail.com>]

Le 08/09/2023 à 16:27, Hesham ElBakoury via Starlink a écrit :
> This MIT Technology Review report provides the details of the work 
> done by Professor Humphrey and his team to reverse engineer starlink 
> signals: 
> https://www.google.com/amp/s/www.technologyreview.com/2022/10/21/1062001/spacex-starlink-signals-reverse-engineered-gps/amp/

Thanks for the pointer.  I have not read it.

But, I would challenge the concept of reverse engineering there to say 
that one might not need to reverse engineer at very low level to achieve 
that effect.

In a WiFi and GSM setting (not starlink), people do not reverse engineer 
it too much to make localisation out of wifi beacons and GSM cell info.  
Just like starlink, WiFi and GSM were not designed to do localisation, 
yet the current smartphone localisation info ('where am I') is relying 
extensively on GSM, WiFi, bluetooth and other non GPS signals.  I 
suspect it relies on these even more than on signals that were designed 
to do localisation like the Galileo and other countries signals.

In the future, with this constant increase of starlink availability, it 
might be possible to add starlink as an additional source of time and 
localisation (in addition to GSM, WiFi etc) without any reverse 
engineering.  Just listen to the signals (if they are not crypted, I 
dont know) and derive some info from there.

The position of the starlink sats themselves is public info.  So, it is 
very possible to derive a more precise the position ('where am I') from 
the sats that I receive, and knowing which sats are currently in view at 
my approximate location.  Probably there are already some smartphone 
apps that show that somehow.

The interesting thing would be the other way around: non-starlink sats 
listen to starlink signals emitted by ground user antennas and tell 
where these users are.   Even that would not require very deep reverse 
engineering, but the achieved effect could be very effective.  Of 
course, it depends how it is used, for which purpose.

Alex

>
> It is worth it to mention that Elon Musk does not like using starlink 
> for navigation which is the motivation behind Humphrey
>
> Although the report mentions Humphrey non-peer reviewed paper, 
> actually the peer-reviewed paper has been published here
> https://ieeexplore.ieee.org/abstract/document/10107477?casa_token=ws8yh0wRnRsAAAAhA:Q0LZ3Gt-vxKsGQ170OhIp4X4tQoa_U2m6DlBKY7a-A0EYxTzRd4VLzpHkZKSglEPiMEJKK4XTHk 
> <https://ieeexplore.ieee.org/abstract/document/10107477?casa_token=ws8yh0wRnRsAAAAA:Q0LZ3Gt-vxKsGQ170OhIp4X4tQoa_U2m6DlBKY7a-A0EYxTzRd4VLzpHkZKSglEPiMEJKK4XTHk>
> Hesham
>
> _______________________________________________
> Starlink mailing list
> Starlink@lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/starlink