RE: Why ?

[EXT Tony Bartoletti <azb@llnl.gov>]

Eric,

"Layers" of location is an interesting notion.  When I first learned of
this forum, I assumed location referred exclusively to the pure spacial
coordinate sense (and wonder, still, about a representative range of use-
cases.)  But for (say) criminal/jurisdictional issues, I am sure that
law enforcement would benefit from broader determinations of location,
such as "the perpetrator was definitely in Alameda County, California
at the time he launched the global cyber-attack from his cell phone."

Of course, knowing the "geometric" location with enough precision
allows one to determine the "jurisdictional" location immediately.
But the latter value is less an intrusion into privacy.  Knowing that
I am speaking from Alameda County is not a great deal of help to a
person who wishes to "stalk" my movements.

I am still quite uncertain of the intended capabilities of the protocol.
In particular, quite "how" the location comes to be (originally) known,
how often it might be updated, etc.

Speaking of uncertainty (Heisenberg's), I don't think that will be much
of an issue for awhile.  Unless you need to know my location to well below
the nanometer scale, or my relative rate of motion is an appreciable
fraction of the speed of light, it needn't be considered.  In contrast,
there will be real (cost?) tradeoffs regarding certainty across (say)
meter vs. kilometer resolutions, I expect.

___tony___
   
At 07:31 AM 01/13/2000 -0600, EXT Eric Crane wrote:
>I have been following along with interest and have these points to make:
>
>With all the tunneling going on today, and in the face of the wireless
>industry's proposals to tunnel even more for mobile IP users, how do we
>determine which IP address in the stack to use for location?  Some of the
>inside IP addresses may be private network addresses and resolve into
>garbage location information.
>
>Could we look at proposing "layers" of location information?  If you look at
>a snail-mail address, there are typically three layers of geographic
>information going from specific to broad.  A location request need not
>necessarily return the completely defined special location down to the
>particle level.
>
>Layers of location information may be extremely useful.  Say an inmate
>restriction service could force intimate knowledge of location to be made
>public (like the new laws in the US making sex offender's address public).
>But other services could provide varying levels of privacy, and a
>prospective customer choose an appropriate level of privacy by service, and
>even within a service.  SO the customer that buys a phone card is virtually
>annonomus, and the customer that signs up for a phone contract must divulge
>a billing address.   An ISP or an e-vendor may need to know the legal
>jurisdiction of a customer in order to abide by taxation laws.  (Or should
>we just have an e-tea party???)  So maybe certain points about location
>could be public knowledge such as country or LATA, etc.  But other more
>explicit information regarding location such as geophysical coordinates may
>be private to the subscriber.
>
>It also occurs to me that the Hiesenburg (sp) uncertainty principle applies
>to determining the location of moving objects, such as mobile users.  It may
>not be possible to exactly locate a target.
>
>Eric Crane
>Motorola, Inc.
>Arlington Heights, IL


Tony Bartoletti (azb@llnl.gov)
Information Operations, Warfare and Assurance (IOWA) Center
Lawrence Livermore National Laboratory