Re: Q: Factoring a number whose Euler phi is known?
From: Stefan Katzenbeisser (katzenb@REMOVEdbai.tuwien.ac.at)
Date: 02/12/03
- Next message: Pete: "Re: attacks on 'RC5'"
- Previous message: Paul Crowley: "Re: OT: Value of Semiconductors Debug Mode Information"
- In reply to: Klaus Pommerening: "Q: Factoring a number whose Euler phi is known?"
- Next in thread: Matthew Johnson: "Re: Q: Factoring a number whose Euler phi is known?"
- Reply: Matthew Johnson: "Re: Q: Factoring a number whose Euler phi is known?"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]
From: Stefan Katzenbeisser <katzenb@REMOVEdbai.tuwien.ac.at> Date: Wed, 12 Feb 2003 15:25:21 +0100
In article <3E4A3582.D00FB8F2@imsd.uni-mainz.de>, Klaus Pommerening wrote:
>
> Remark. There is a probabilistic algorithm: Choose a "public" exponent
> e,
> construct a corresponding "private" exponent d such that ed \equiv 1
> \pmod {\phi(n)}. Then use d to factor n by the well known probabilistic
> algorithm.
There is a far better algorithm for that problem: try to
find a non-trivial square root of 1 modulo n, given
an "oracle" that computes any multiple of d=\lambda(n),
where \lambda is Carmichael's function (\phi is always a multiple
of \lambda). Any such square root immediately yields to a
factorization of n.
If d=(2^k)d', there exists a probabilistic algorithm that
outputs such a non-trivial square root in time polynomial
in k and log(n). If n is a RSA modulus, the success probability
is greater 1/2, but the algorithm should also work for arbitrary n
(however, I don't know its success probability right now).
>
> My question asks for a deterministic algorithm.
To my knowledge, there is no deterministic algorithm
known, but I may have missed something here. It _is_ known
that there is a deterministic algo for that problem once
you assume the correctness of Riemann's hypothesis, though...
Hope this helps!
-S
- Next message: Pete: "Re: attacks on 'RC5'"
- Previous message: Paul Crowley: "Re: OT: Value of Semiconductors Debug Mode Information"
- In reply to: Klaus Pommerening: "Q: Factoring a number whose Euler phi is known?"
- Next in thread: Matthew Johnson: "Re: Q: Factoring a number whose Euler phi is known?"
- Reply: Matthew Johnson: "Re: Q: Factoring a number whose Euler phi is known?"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]
Relevant Pages
|