Re: The Onion on Lava Lamps

From: Tony Finch (dot@dotat.at)
Date: Tue Feb 20 2001 - 11:05:29 PST


Jeff Bone <jbone@jump.net> wrote:
>
> Some guy at Hahvahd is poking at this, too.

>From the IP list:

http://www.nytimes.com/2001/02/20/science/20CODE.html

        A computer science professor at Harvard says he has found a
        way to send coded messages that cannot be deciphered, even by
        an all-powerful adversary with unlimited computing power. And,
        he says, he can prove it. [...]

        In essence, the researcher, Dr. Michael Rabin and his Ph.D.
        student Yan Zong Bing, have discovered a way to make a code
        based on a key that vanishes even as it is used. While they
        are not the first to have thought of such an idea, Dr. Rabin
        says that never before has anyone been able to make it both
        workable and to prove mathematically that the code cannot be
        broken. [...]

        For now, Dr. Rabin's idea is simply a scheme backed up by a
        mathematical proof that he has been presenting to scientists
        at seminars. No company is lurking in the background to sell
        it, and Dr. Rabin says he has no commercial interests in it.

        "I never commercialize anything," Dr. Rabin said. "I am not in
        that business." Instead, he said, he did the work because it
        was a challenge. [...]

        The coding starts with a continuously generated string of
        random numbers, say from a satellite put up to broadcast them
        or from some other source. The numbers can be coming by at an
        enormous speed - 10 million million per second, for example.

        The sender of a message and its recipient agree to start
        plucking a sequence of numbers from that string. They may
        agree, for example, to send a message, encoded with any of
        today's publicly available encryption systems saying "start"
        and giving instructions on capturing certain of the random
        numbers. As they capture the numbers, the sender uses them to
        encode a message, and the recipient uses the numbers to decode
        it.

        An eavesdropper can know the mathematical formula used to
        encode and decode, but without knowing the exact sequence of
        random numbers that were used in the formula to send a
        particular message, the eavesdropper cannot decode the
        message. And the only way to have that sequence is to just
        happen to be storing numbers from the unending stream at
        exactly the right moment.

        If the eavesdropper, for example, had a secret way to decode
        the message saying "start" and it took a minute to do the
        calculation needed to decode it, it would be too late by the
        time the eavesdropper got going. The sender and recipient
        would already have their string of numbers and that string of
        numbers, once broadcast, could never be retrieved. It would be
        infeasible to store the endless string of numbers in any
        computer and so they are essentially gone forever.

        Often, Dr. Rabin said, eavesdroppers will capture and store
        encoded messages hoping to decode them at later, either when
        computers have improved - making it easier to do the
        calculations to break a code - or when the method for encoding
        and decoding is known, perhaps because it has been stolen.
        But, he said, messages encoded with his system can never be
        broken by these means because the random numbers used in
        encoding and decoding are used once and are never stored. [...]

[presumably the recipient of the message has to cache the random
numbers for a time in order to decode the message, in which case that
cache is a weak point; the article assumes the cache doesn't exist]

Tony.

-- 
f.a.n.finch    fanf@covalent.net    dot@dotat.at
VIKING NORTH UTSIRE SOUTH UTSIRE NORTHEAST FORTIES: WEST VEERING NORTHWEST 5
TO 7, PERHAPS GALE 8 LATER. RAIN THEN SHOWERS. MODERATE BECOMING GOOD.



This archive was generated by hypermail 2b29 : Fri Apr 27 2001 - 23:18:00 PDT