1994-12-12 - Re: Real-time surveillance of the police

Header Data

From: tcmay@netcom.com (Timothy C. May)
To: sandfort@crl.com (Sandy Sandfort)
Message Hash: bd32f9f97013edfc0bac74004331ef4285604d616d5567332a56960df4cc2b11
Message ID: <199412122252.OAA05139@netcom19.netcom.com>
Reply To: <Pine.SUN.3.91.941212125649.28598A-100000@crl.crl.com>
UTC Datetime: 1994-12-12 22:53:55 UTC
Raw Date: Mon, 12 Dec 94 14:53:55 PST

Raw message

From: tcmay@netcom.com (Timothy C. May)
Date: Mon, 12 Dec 94 14:53:55 PST
To: sandfort@crl.com (Sandy Sandfort)
Subject: Re: Real-time surveillance of the police
In-Reply-To: <Pine.SUN.3.91.941212125649.28598A-100000@crl.crl.com>
Message-ID: <199412122252.OAA05139@netcom19.netcom.com>
MIME-Version: 1.0
Content-Type: text/plain


Topics I'll discuss: surveillance tapes of cops, localizer bandwidth
needs, and digital timestamping.

There's _some_ Cypherpunks relevance here, as this thread deals with
the issue of surveillance, self-surveillance, escrowed records, etc. 

I have another point of view to add: I don't want to see my local cops
have all of their actions videotaped and reviewed for quite another
reason--I don't want a "mechanization" of the enforcement process!

Cops, for all of their faults, also have some positive personal
characteristics: they use discretion in enforcing laws, they let folks
off with warnings or shrugs, etc. I shudder to think about a world in
which the hideously complicated legal code is enforced by cops who
have video cameras mounted behind them to verify that they never gave
a sucker^H^H^Hcitizen-unit a break. 

(I am slightly reluctant to tell Sandy I have some good opinions of
the California Highway Patrol, but I do. I elected to take a one day
class to get a speeding ticket taken off my record, and the class was
taught by an extremely able, pragmatic retired CHP officer. He had all
kinds of tips about driving, accidents, etc., having seen 500 or more
fatalities in his 25+ years with the CHP. He also talked about the
local judgment, or discretion, that CHP officers have to have. For
example, to raise tax revenues, California ruled that "fix-it tickets"
(no fee, no fine, just present evidence that a minor automobile flaw
has been corrected) would no longer be allowed, that even the most
minor infractions--broken tail-light, cracked windshield, etc.--would
have to go through the legal process, with special soak-the-driver
"administrative" (really, revenue enhancement) fees. So the CHP is
simply not playing along, and they've found a way to avoid the process
by issuing a different kind of ticket that is in fact a fix-it
ticket. He also gave other examples that left me with a more favorable
impression of cops....perhaps this was his intent?)

(The longer-than-I'd-planned section above is meant to show that even
a cop-disliking skeptic like me can be convinced that some good is
done is by them, and that not all cops are bad. Perhaps my on-day
exposure co-opted me the way certain folks in D.C. got co-opted?)

Anyway, I don't _want_ a mechanistic enforcement of *all* laws, with a
video camera second-guessing the cop. I don't want 43,761 laws being
enforced religiously. (If we had just the 13 or 14 laws I think we
need, strict enforcement would be a good thing, but not when the
43,748 extra laws are cluttering up the books.)

"Officer, the video records of 1994-12-12 14:22:31 PST indicate your
patrol unit was within visual range of a 324.986.666 ("Bicycle Helmet
of Improper Color") and yet you ignored this misdemeanor. This being
your 3rd such oversight this month, you are being reassigned to the
Discipline Platoon at Camp Pendleton. Dismissed."

Forcing the cops to surveill themselves will almost certainly take
away what enforcement discretion they now have.

Sandy Sandfort wrote:

(quoting someone else)

>     ObCrypto/Privacy:  I suspect there would be an immense
>     amount of radio traffic involved in keeping track of a
>     substantial group of people ... [I doubt] the ability of
>     the receivers to digest it all in real time.
> 
> (1)  My suggestion was for (probably local) recording, not real
> time monitoring of video; therefore, no bandwidth problems.
> 
> (2)  Tim might want to comment, but my understanding of the
> localizer technology is that it too is somewhat "local" and that
> polling or burst transmission keeps the bandwidth requirements at
> manageable levels.

On this bandwidth point, it is certainly true that N
localizers/transceivers communicating at M bits per second with some
maximum carrier frequency can "overload" the "free space channel."
(Crudely, when N x M > .5 B, where B is the "bandwidth" of the
communications channel.)

For example, a million people each trying to communicate a thousand
bits per second would imply an aggregate of a billion bits pers
second, barely possible if the carrier frequency maxes out at a few
gigahertz. This is Shannon's Theorem, of course. And this is within
"one space," nearby. In the real world, with a few miles as the
effective range, the "crowding" is not severe. (I'm ignoring other
users, radio and t.v., cellular, RF noise, etc. Important things to
consider in a more detailed calculation.)

Ten thousand such units, in a space a few miles square, each "trying"
to communicate 1000 bits per second, would result in only about 10^7
bits per second, aggregate. This is far, far below any Shannon limits
(of course, there are other users of the spectrum who may "step on"
the users here...these are all tradeoffs to consider, and that _have_
been considered). Fortunately, localization doesn't typically need
1000 bits per second, in any case.

(A skier lost in the mountains, a child who has wandered away....all
are cases where the needed bits per second is _much_ less than 1000
bps. Indeed, most of the localizers are either not sending info at
all, or very sporadically.)


Finally, the problems of "proving" the surveillance or audio/video
records are not later changed by cops or others is easily solved--by
something of great crypto relevance.

Namely, digital time-stamping, a la Haber and Stornetta. No "trusted
third party storage" is needed: just hash the surveillance records a
la Haber and Stornetta and the resulting published hash cannot at a
later time be forged. (I devoted much space to this in the FAQ, so
grep it for this if interested.)

--Tim May

-- 
..........................................................................
Timothy C. May         | Crypto Anarchy: encryption, digital money,  
tcmay@netcom.com       | anonymous networks, digital pseudonyms, zero
408-688-5409           | knowledge, reputations, information markets, 
W.A.S.T.E.: Aptos, CA  | black markets, collapse of governments.
Higher Power: 2^859433 | Public Key: PGP and MailSafe available.
Cypherpunks list: majordomo@toad.com with body message of only: 
subscribe cypherpunks. FAQ available at ftp.netcom.com in pub/tc/tcmay






Thread