Here‘s a new one …
Gps spoofing raises alarms
As mentioned in the article this seems to be more than just GPS spoofing. It will be interesting to see what causes the INS system and VOR/DME to also fail. Sounds almost like a bug in specific avionics systems if GPS clock is lost.
chflyer wrote:
It will be interesting to see what causes the INS system and VOR/DME to also fail.
Where is this written in the article? FMS is supposed to take different data sources in calculating position and estimating its accuracy based on Kalman filter calculations. I don’t see how sudden loss of GPS signal or sudden change in GPS provided location can confuse FMS which has IRS and DME/DME (and VOR/DME and VOR/VOR) inputs.
Emir wrote:
Where is this written in the article? FMS is supposed to take different data sources in calculating position and estimating its accuracy based on Kalman filter calculations. I don’t see how sudden loss of GPS signal or sudden change in GPS provided location can confuse FMS which has IRS and DME/DME (and VOR/DME and VOR/VOR) inputs.
“Supposed” is the key word. I would not have too high expectations about FMS being bug-free (see https://euroga.org/forums/hangar-talk/15093-uk-airspace-closed for example of a bug in aviation-related sw). Especially on a computation path (missing/strangely/partially corrupted data) that is not properly exercised in normal use. You can test a gazzillion different combinations of corrupted data and still miss the case that gets you. You can also have a provably-correct high level algorithm, only to be taken down by an implementation detail that subtly invalidates an assumption on which the algorithm’s correctness is based on.
Emir wrote:
Where is this written in the article?
The last sentence in the first paragraph. Ok, it doesn’t say that the INS fails, rather that it becomes unusable…. but that amounts to the same thing.
esteban wrote:
I would not have too high expectations about FMS being bug-free (see https://euroga.org/forums/hangar-talk/15093-uk-airspace-closed for example of a bug in aviation-related sw). Especially on a computation path (missing/strangely/partially corrupted data) that is not properly exercised in normal use. You can test a gazzillion different combinations of corrupted data and still miss the case that gets you. You can also have a provably-correct high level algorithm, only to be taken down by an implementation detail that subtly invalidates an assumption on which the algorithm’s correctness is based on.
I can accept that there can be bugs in any software but FMS and flight plans processing software are absolutely in different categories. Moreover the article lists several different types (Boeing, Gulfstream, Dassault and Bombardier aircraft) so the probability that all have same bug processing failed inputs is virtually zero.
chflyer wrote:
The last sentence in the first paragraph. Ok, it doesn’t say that the INS fails, rather that it becomes unusable…. but that amounts to the same thing.
It’s added in brackets which means that the author of the article concluded about this reference from reports. I would be very sceptical about this stating INS rather than IRS. However, anything is possible…
Emir wrote:
I can accept that there can be bugs in any software but FMS and flight plans processing software are absolutely in different categories. Moreover the article lists several different types (Boeing, Gulfstream, Dassault and Bombardier aircraft) so the probability that all have same bug processing failed inputs is virtually zero.
Quoting: this is not ordinary GPS jamming, which is a common occurrence in the area. The attacks have all been the same so far. The specific GPS receiver on a single aircraft is sent a signal that shifts the displayed position by 60 nautical miles. The aircraft’s nav systems freak out at the sudden change in data and in almost all cases the screens become useless.
I can imagine the programmers of FMS correctly implemented handling missing GPS data … but handling missing data is much easier than handling Byzantine data that are adversarially crafted to break the nav systems (e.g. by Iranians trying to learn how to play games with US drones). That might not have been in the specs for the FMS’s … and even if so, it is easy to overlook some of the possible attack approaches.
Has anyone actually read the linked article?
is reporting that aircraft using Airway UMB688 in northern Iraq are experiencing complete navigation system failures because the hacker replaces the position data beamed by the GPS signals with false coordinates
N Iraq. No Sh1t Sherlock. Yeah, we fly there all the time…
A linkage to other nav sources is interesting though, but not surprising. Like that INS which crashed when crossing the 180 degree longitude. It’s called “software” 
The spoofing is probably done well (military grade) and since there is no SBAS coverage, you get no verification of the GPS solution.
I would be very very very much surprised +if US military drones cared much about GPS; I mean, everyone and their dog can now afford to launch their own constellation of satellites, why wouldn’t the US military have one or two. For redundancy, you know.
US mil uses GPS (a separate encrypted band) but a jammer has to be above the receiver altitude to work effectively. That way, guided weapons flying at altitude can have GPS guidance along most of the trajectory. And they have INS for the rest of the time.
