GPS interference testing

When you call for Notams, as I understand it you must
ask specifically if there is anything slated for GPS
testing, otherwise you will not be told what's up.
For those who want to see what can happen please look
at my flght log for 16 May on OLC. The test not only
screwed up the gps instrument, it also 'moved me' on
my PDA into the White Sands SUA.
I am still waiting for a response from OLC as to the
best way to correct the flight claim.
howard banks/1XX

"Howard Banks" wrote in message
...

For those who want to see what can happen please look
at my flght log for 16 May on OLC. The test not only
screwed up the gps instrument, it also 'moved me' on
my PDA into the White Sands SUA.

Dunno, Howie, it looks like a Pez-type alien abduction to me. Where WERE
you those 14 minutes?

Bill Daniels

With the recent concerns (last year) over the security of traces, I'm
surprised the validation softward didn't have some trouble with your
data. If I were an alogorithm, I would take special interest in your
118 statute mile leg completed at 474 miles per hour. Fly impressive do
you, Jedi master.

We've seen similar problems in the vicinity of you know where. We'll
sometimes get bizarre navigation solutions, but we're yet to see any
wormholing of traces.

OC

Howard Banks wrote:
When you call for Notams, as I understand it you must
ask specifically if there is anything slated for GPS
testing, otherwise you will not be told what's up.
For those who want to see what can happen please look
at my flght log for 16 May on OLC. The test not only
screwed up the gps instrument, it also 'moved me' on
my PDA into the White Sands SUA.
I am still waiting for a response from OLC as to the
best way to correct the flight claim.
howard banks/1XX

wrote:
With the recent concerns (last year) over the security of traces, I'm
surprised the validation softward didn't have some trouble with your
data. If I were an alogorithm, I would take special interest in your
118 statute mile leg completed at 474 miles per hour. Fly impressive do
you, Jedi master.

The algorithm that checks for this kind of problem resides in the head
of the official evaluating the flight. Clearly bad points, temporarily
dropped GPS signals or power loss, etc, are infrequent and best handled
by a person, as they do not automatically invalidate a file or a flight.
The IGC validation software checks the flight file for (basically)
tampering only, and not external problems.


--
Change "netto" to "net" to email me directly

Eric Greenwell
Washington State
USA

Good point.

But what if it was just a few dozen extra miles at 150mph? Enough, for
example, to get a record, but not so much that the eyes go wide with
wonder?

Seems clear that tampering can be managed from outside the unit while
producing secure, valid record files. I've been amused, for instance,
that the OLC will highlight a flight if it is downloaded improperly or
transfered among media prior to uploading, but otherwise assume all is
well.

It's tough to thwart a determined cheater. But since the vast majority
of cheating is a matter more of convenience than maliceaforethought, to
most of us the emphasis on security seems draconian. Now that we have
clear evidence that there are external devices that can spoof our
recorders, is it the next step to require that recorders have some
protection against such possibilities? Or is this where we throw up our
hands and say "enough" and put the responsibility back into the hands
of the observer?

As an SN10 owner, I've always wondered at the prohibition of external
GPS engines. Doesn't seem like the cable is really at issue anymore.
Granted, it would take a smart guy with access to the proper tools, but
that's also true of altering files post flight.

Note that 1XX's trace may not be the result of a random anomaly but an
indication of intential interference from a ground based transmitter.
If I can spoof the signal, can't I then create the file? With the right
set of tools and expertise (and an observer willing to cut some
corners), I could fly to 100,000 feet and circumnavigate the globe...
and deliver uncorrupted proof of my accomplishment. Getting people to
accept the proof then is just a matter of proportion.


Eric Greenwell wrote:
wrote:
With the recent concerns (last year) over the security of traces, I'm
surprised the validation softward didn't have some trouble with your
data. If I were an alogorithm, I would take special interest in your
118 statute mile leg completed at 474 miles per hour. Fly impressive do
you, Jedi master.

The algorithm that checks for this kind of problem resides in the head
of the official evaluating the flight. Clearly bad points, temporarily
dropped GPS signals or power loss, etc, are infrequent and best handled
by a person, as they do not automatically invalidate a file or a flight.
The IGC validation software checks the flight file for (basically)
tampering only, and not external problems.


--
Change "netto" to "net" to email me directly

Eric Greenwell
Washington State
USA

wrote:

Seems clear that tampering can be managed from outside the unit while
producing secure, valid record files. I've been amused, for instance,
that the OLC will highlight a flight if it is downloaded improperly

This is a problem.
or transfered among media prior to uploading,

Transferring among media is not a problem. The file and it's signature
are preserved when the file is copied between hard drives, memory cards,
floppy disks, sent by email, or via the web. Consider: I download my 302
file to my Ipaq memory card, the file is transferred via Activesynch
(desktop) or memory card reader (laptop) to a computer hard drive, then
uploaded to the OLC.

but otherwise assume all is well.

There is a procedure for disputing flight claims, but the OLC does not
routinely attempt to verify flights. Quite reasonable, considering the
goal of the OLC.

It's tough to thwart a determined cheater. But since the vast majority
of cheating is a matter more of convenience than maliceaforethought, to
most of us the emphasis on security seems draconian. Now that we have
clear evidence that there are external devices that can spoof our
recorders,

This isn't news, but has been discussed here and elsewhere for years.

is it the next step to require that recorders have some
protection against such possibilities? Or is this where we throw up our
hands and say "enough" and put the responsibility back into the hands
of the observer?

I think we did this quite a while ago. It's always been known that the
recorders don't offer protection against external signals, either
intentional or unintentional.

--
Change "netto" to "net" to email me directly

Eric Greenwell
Washington State
USA

wrote:

Seems clear that tampering can be managed from outside the unit while
producing secure, valid record files

snip

If I can spoof the signal, can't I then create the file? With the
right set of tools and expertise (and an observer willing to cut some
corners), I could fly to 100,000 feet and circumnavigate the globe...
and deliver uncorrupted proof of my accomplishment. Getting people to
accept the proof then is just a matter of proportion.

That same log file included a barometric altitude trace. During the
period of interference it diverged from the gps altitude by an obvious
amount.

If you want to produce a valid igc file depicting a plausible world
record, you will have to do a lot of work to get a the time, barometric
altitude, gps altitude and gps co-ordinates to match with each other to
a standard where it will stand up to some scrutiny.

The sad thing is I think that this has just dealt a blow to those, who
like myself, who would like to see records from commercial grade GPS
equipment accepted for badge claims (especially silver and gold).


Ian

"Ian Forbes" wrote in message
...
wrote:

The sad thing is I think that this has just dealt a blow to those, who
like myself, who would like to see records from commercial grade GPS
equipment accepted for badge claims (especially silver and gold).


Ian

Ian,

I'm afraid you may be right, but that doesn't mean it's logical. Let's
take a journey down memory lane for a moment:

Suppose we are back in the dark ages, when Barographs ruled the sky. Then,
let's suppose I really, really wanted my gold climb but lacked either the
aptitude or patience to get it. What to do? For starters, I could take a
stroll to my local clock repair shop. The mechanism on the inside of a
typical barograph is no more sophisticated (actually less sophisticated)
than your typical grandfather clock (this isn't just supposition - a club
member had his Winter repaired by a local Cuckoo Clock maker). A little
adjustment to the "gain" on the recording stylus with a commensurate tweak
to the rate of rotation (so as not to raise any suspicions about the rate of
climb) and voila - the extra thousand or so feet I need. Of course, this
is all done on a recently calibrated device, complete with current
calibration trace.

Is this far-fetched? Absolutely. But no more farfetched than the arguments
being put forth by those trying to find reasons not to allow COTS GPS
devices for basic badges. The sport survived for 70+ years without dual
recording mechanisms, and we still seemed to be confident in the validity of
the majority of flight claims. Why have we lost all sense of proportion?

Let's fast forward to today. Suppose I want to set the Standard Class
straight distance to a goal record (okay, not an FAI record, but bear with
me). I mount my secure logger to the glider, declare my flight, and off I
go. Oh, but I decided to strap the 18M tips on, just to give me that 10%
extra float to extend the last glide. Could I find an OO willing to look
the other way (or who might be ignorant enough not to notice an extra 3M of
span)? I'm sure I could, especially if I wanted to make a small donation.
So, off I go, and I finish the flight. The log file is downloaded in the
presence of a bonded agent of Price Waterhouse Coopers, the 512bit
encryption checks out, and I'm now the record holder. My East Coast buddies
know me as an outstanding standard class pilot :-)), so no red flags are
raised.

Is this far-fetched. Absolutely. But is it more likely than somebody
spending a few hundred hours and $$ building a GPS-signal generator. I
think so. The first rule of security is to look at the system as a whole.
Financial services companies learned long ago that all of the firewalls in
the world can't prevent a couple of individuals on the inside from doing a
world of harm.

The only solution to the intransigence of certain parties (I have agreed to
refrain from name calling) is for the grass roots folks to go directly to
their national Soaring Associations and say - "enough". As long as there
is a role for an OO in the system (and there always will be), security is
simply a matter of good process. It will never be perfect.

Erik Mann (P3)