[time-nuts] GPS Spoofing

[Jim Lux]

On 7/26/13 8:45 PM, J. Forster wrote:
> I gather from the article, the GPS position was spoofed and the autopilot,
> in bringing it back to where it was supposed to be, actually took it off
> course.
>
> There are places where a few hundred feet makes a big difference, viz. the
> Costa Concordia.
>
> IMO, this is a very convincing reason for something like LORAN.

I think it's a convincing argument for a captain who pays attention to 
the other navigation instruments and doesn't blindly follow the GPS.

It's also a convincing argument that shipboard 
automation/autopilot/autocontrol vendors need to make more sophisticated 
software (which I suspect they do, particularly on 200+ foot ships.. I 
would imagine that there are some aspects of this demo that are 
contrived.)  The ship making and driving business is pretty unregulated. 
It's all about what the owner of the ship is willing to pay (or what he 
needs to get liability insurance, if he wants).  There's nothing even 
remotely like DO-178 for shipboard stuff.

The folks doing stabilized oil rigs probably have sophisticated systems, 
but they're also using IMUs and other stuff. Ditto for high value things 
(oil tankers, warships).  Molasses tankers? They're probably lucky to 
have a functioning compass and some old charts.


I'm not sure, though, that looking at the big picture, whether your tax 
dollars are better spent on LORAN, or on some other precision navigation 
method or on making jam resistant GPS receivers (which do, in fact 
exist, and make use of things like direction of arrival of the signal..)

Note that a GPS system with 3 antennas (as is common in systems that use 
GPS to derive attitude/orientation) would be extremely difficult to 
spoof, and would be VERY inexpensive to implement.  Either the carrier 
phases and code phases are consistent for all the received signals or 
they're not.  A jamming signal coming from the wrong direction will not 
have the right direction of arrival relative to the platform 
orientation.  One wrong signal might be tolerable (multipath, etc.) but 
with a multi satellite fix, I suspect it would be hard to do it.

Sure, one could throw up N pseudolites on a bunch of UAVs, etc., but 
that's getting to be a bit noticeable.


For what it's worth, I don't know that LORAN has the performance to 
avoid a Costa Concordia type foul up (assuming they were crazy enough to 
do the near pass in the fog, so visual navigation didn't work)

I seem to recall that LORAN had 1/4 nmi kinds of accuracy.  it would get 
you to the channel or mouth of the harbor, but not get you into your 
berth. You might be familiar with the local propagation anomalies and 
get better accuracy with experience in your local waters.






>
> -John
>
> =================
>
>
>
>> I boat?  The backup is a competent captain.  He'd see the compass heading
>> move and quickly disengage the autopilot.   I had a boat for years  I'd
>> notice a 5 degree change.  Mine was a sailboat so I'd be more sensitive to
>> heading changes than a power boater but still the human is the backup.
>>
>> Most autopilots don't directly follow GPS, they use GPS to determine a
>> heading, follow it then use GPS to detect drift and re-compute the
>> heading.
>>   the heading would be held by a compass sensor in a low-cost setup or in a
>> larger setup a lazer ring gyro backed up by a compass.     So a spoofed
>> GPS
>> would cause the autopilot to "think" there was a bigger crooswnd or
>> current
>> and make a bigger heading change.
>>
>> I bet you could hijack a drone not a manned vehicle the pilot is trained
>> to
>> monitor the automation and he'd very quickly turn it off thinking it was
>> broken.
>>
>>
>>
>>
>>
>>
>> On Fri, Jul 26, 2013 at 8:41 AM, J. Forster <jfor at quikus.com> wrote:
>>
>>> Prof. Humphry from Texas just reported being able to spoof GPS in the
>>> Med
>>> and take over the nav system of a luxury yacht. He's done this before
>>> with
>>> a drone in the US.
>>>
>>> LORAN as a backup, at least?
>>>
>>> -John
>>>
>>> ==============
>>>
>>>
>>>
>>> _______________________________________________
>>> time-nuts mailing list -- time-nuts at febo.com
>>> To unsubscribe, go to
>>> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>>> and follow the instructions there.
>>>
>>
>>
>>
>> --
>>
>> Chris Albertson
>> Redondo Beach, California
>>
>
>
> _______________________________________________
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.
>