[Cryptography] A software for combining text files to obtain high quality pseudo-random sequences in practice

Jerry Leichter leichter at lrw.com
Wed Jul 12 06:32:33 EDT 2017


> The original paper is by Ueli M. Maurer
> "Conditionally-Perfect Secrecy and a Provably-Secure Randomized Cipher"
Thanks.  That was what I was thinking of (though I don't recall ever reading the original paper, just saw a reference to the idea).

> But there is no mention of the moon-image.
> 
> This reference is made in Schneiers criticism:
> 
> https://www.schneier.com/academic/paperfiles/paper-maurer-stream.pdf
For completeness, from the abstract:  "We should point out that our analysis does not apply to the original cipher proposed by Maurer, where the pool is chosen to be large enough that it is infeasible to read the entire contents of the pool. Unfortunately, Maurer’s cipher is not very practical—in his paper, he proposed digitizing the surface of the moon as one means of getting enough public randomness to make the cipher work—and hence not suited to present-technology implementation. We attack a simplified version, one which is more practical."

Almost 40 years later, there probably do exist databases large enough to make the original idea plausible.  Just consider the concatenation of all Google Street Map images, for example.  Whether there are any practical applications is another question.  (Whether anyone other than Google itself would be able to reliably have access to a fixed set of those images is part of the "practicality" issue.)

                                                        -- Jerry



More information about the cryptography mailing list