[time-nuts] More GPS troubles

[David]

On Thu, 17 Jan 2013 07:08:20 -0800, Jim Lux <jimlux at earthlink.net>
wrote:

>On 1/17/13 6:22 AM, Bob Camp wrote:
>> Hi
>>
>> Most cheap GPS's these days have user friendly firmware update capability.
>> That's been true for quite a while. I'd be amazed if the higher end stuff
>> didn't make updates an easy thing. Bugs in GPS code are not exactly
>> uncommon.
>>
>> The real issue is the need to get the GPS code patched for this kind of
>> thing. Without nutty articles that alarm the marketing department, the work
>> will never get any sort of priority.
>
>But is there *really* a need to patch for this..(as opposed to the other 
>bugs that are on the list)
>
>Consider you're using GPS as a time reference in a "mission critical" 
>application.  You've already got to have some holdover capability.  Do 
>you think someone could set up one of these jammers, jam your GPS, and 
>keep doing it for long enough to extend you past your max holdover time?
>If you're that critical, you need geographically dispersed receivers 
>anyway (struck by lightning?)

Two of the attack examples disabled the GPS receivers, one rebooted
continuously until a hard reset to erase the nonvolatile memory and
the other permanently returned invalid date and timing information,
even after the short term spoofing signal was removed.  Only minutes
or tens of minutes were needed for the attacks.

>In the example of the high accuracy reference networks.. they're 
>*networks* and they are designed so that individual nodes can fail. 
>Spoofing attacks, by their nature, can really only attack one receiver 
>at a time even if the spoof signal covers a long distance(because only 
>for the chosen victim do all the signals line up just right). So now 
>you're talking about N jammers for N receivers, and the cost of your 
>jamming attack is rising.

These low level spoofing attacks involved changing the satellite
ephemeris, satellite almanac, or GPS epoch so they would work against
multiple receivers simultaneously.

>Someone who wanted to deny the use of the network would be better served 
>by sending someone out with aluminum foil/paint to cover the radome of 
>the stations. Or a cutting torch and sawzall. (tougher to do and a bit 
>more obvious than commanding an array of jamming systems from your 
>underground lair)

That would be cheaper at least for now.