home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: sci.math
- Path: sparky!uunet!stanford.edu!CSD-NewsHost.Stanford.EDU!Sunburn.Stanford.EDU!pratt
- From: pratt@Sunburn.Stanford.EDU (Vaughan R. Pratt)
- Subject: Re: ok, i'll bite.
- Message-ID: <1992Nov16.190432.25281@CSD-NewsHost.Stanford.EDU>
- Sender: news@CSD-NewsHost.Stanford.EDU
- Organization: Computer Science Department, Stanford University.
- References: <92320.214859RVESTERM@vma.cc.nd.edu>
- Date: Mon, 16 Nov 1992 19:04:32 GMT
- Lines: 40
-
- In article <92320.214859RVESTERM@vma.cc.nd.edu> <RVESTERM@vma.cc.nd.edu> writes:
- >i understand "cryptography". i understand "np". someone please
- >explain the connection.
-
- NP is the class of problems whose instances encrypt or hide their
- easily seen solution. For example the set of composites is in NP:
- hidden in every composite pq is a factor between 2 and pq-1, an easily
- recognized key giving away the secret of pq's compositeness.
-
- Cryptography is not defined as formally as NP, so one cannot hope for
- an exact correspondence. For example the one-time-pad cryptosystem
- does not have a corresponding NP notion. Here one makes up a random
- bitstring R of length that of the message M, and sends R and R+M (their
- XOR) by two different routes. Each of R and R+M by themselves looks
- completely random, but R+(R+M) = M. Even exponentially much work
- cannot extract M from either R or R+M by itself, showing that the full
- resources of cryptography go beyond mere complexity obstacles of the
- kind posed by problems in NP-P.
-
- (Around 1978 Adi Shamir found a neat generalization of this to n
- messages sent by that many routes, such that one must intercept at
- least m of them to read the message. If m = n/2 the adversary must
- catch at least half the couriers to read the message, and block at
- least half to stop it getting through. With only m-1 couriers the
- message remains completely hidden. Other applications: shifty-eyed
- couriers, infiltrated cells, etc. Technique: make the parts of the
- message the values of a degree m-1 polynomial at 1,2,...,n where the
- message itself is the value at 0.)
-
- However as a corollary of plausible assumptions about channels,
- computing time, and application details, many practical cryptosystems
- turn out to be no harder to crack than to decide membership for some
- set in NP. Cracking RSA for example reduces to factoring; it is in NP
- (but not known to be NP-hard) to decide whether the i-th bit of the
- larger factor of pq in the published key is 1. (The converse
- reduction, from factoring to cracking, is not known for RSA, but it is
- known for a variant of RSA due to Michael Rabin, which is as hard to
- crack as to factor integers.)
- --
- Vaughan Pratt A fallacy is worth a thousand steps.
-