From: David Lesher <wb8foz@nrk.com>
Date: Sun, 14 Apr 1996 04:11:29 +0800
To: cypherpunks@toad.com (Cypherpunks)
Subject: Re: No matter where you go, there they are.
Message-ID: <199604131648.MAA08425@nrk.com>
MIME-Version: 1.0
Content-Type: text


I asked a friend charged with spending money in this area.
Here's his reactions, and those of someone he consulted:


==================
..There is a business case flaw in the proposal. The scheme as laid
out only provides a high confidence factor on the users location 
(+- 10M), NOT on the actual identity of the user.  Thus it would
have to be coupled with some other form of I&A.  The business
question then arises, is the confidence delta provided by this
scheme consistent with the cost delta?  My first take is that the
answer is no.

===================
Re spoofing the system.  If you think about the geometry of the
problem, the delay to be induced for each satellite is a time
varying function of the satellite's position, the reference site,
and the target. It will vary from positive to negative values for
many satellite passes.  It can be precomputed, but the precomputed
adjustment will be in error by some amount due to the orbital
perturbations mentioned in the original article.  The most common
prediction error is the in-track postion, which inconveniently for
the predictor has the greatest effect on one's ability to calculate
the right delays.  Anyone with access to a decent orbital prediction
code and access to the statistics on orbital perturbations should be
able to calculate the expected delay prediction error as a result
of the orbital uncertainties.  If the delay prediction errors are
detectable, then it seems the system is secure after all, allthough
other reservations on operational suitability might still apply.

Note that the sensitivity is a function of the distance between the
reference site and the target.  If you could collect your reference
signal across the street from the target, it would make spoofing
a lot easier.

-- 
A host is a host from coast to coast.................wb8foz@nrk.com
& no one will talk to a host that's close........[v].(301) 56-LINUX
Unless the host (that isn't close).........................pob 1433
is busy, hung or dead....................................20915-1433