Re: U-Boat Enigma procedure(Long)

From: Jim Gillogly (jim_at_acm.org)
Date: 07/26/03 

Date: Sat, 26 Jul 2003 19:13:59 GMT

Dan Girard wrote:
> I'm curious as to just how much better the upgraded version performs
> compared to the version described in your paper. How much text would
> each have to have in order to have a good chance of solving a 3-rotor
> message with ten plugboard connections? For a message as long as the
> maximum allowed in the German navy, how many plugs could each version
> handle and still have a good chance of success?

I haven't checked with the kind of rigor I used in my paper. The
main improvement is that it does a generalized hill-climbing across
the range of settings, rather than focussing on one setting at a time.
If it reaches a local maximum, it starts again from another random
starting point -- what I call "shotgun hillclimbing". It's an
improvement because it's a strict superset of the work I described.
In a sense it's less interesting, because it uses more computational
power than the hottest machine of the day (which I assume was the
Zuse Z-3) had available.

I ought to run the data you suggest, though. After a number of other
projects!

> Could the updated version solve a four-rotor message if the rotor
> order and ring-settings were known, but not the plugs or the message
> key? I'm wondering how I would proceed if my program were to give me
> a "possible solution" that started out something like
> "VONVONBDUUUXOFFIZIERDORA..." and then became unintelligible (highly
> unlikely, but not impossible).

Anything that's known is a big help, and narrowing down these two
variables would definitely give a leg up. Again, I've tried it only
on a few data points. 

-- 
	Jim Gillogly
	Mersday, 3 Wedmath S.R. 2003, 19:09
	12.19.10.8.2, 11 Ik 10 Xul, Ninth Lord of Night

Relevant Pages

  • Re: U-Boat Enigma procedure(Long)
    ... >> messages, with ten stecker connections, would be vulnerable to it? ... > under discussion with no success. ... each have to have in order to have a good chance of solving a 3-rotor ... maximum allowed in the German navy, how many plugs could each version ...
    (sci.crypt)
  • Re: HELP PLEEEEASE!
    ... > I have even tried hooking another stero up and to no avail again. ... >> Ok, I though by tape you was doing vcr tapes, but where your mic plugs ... >> the computer you should have two other plugs, ...
    (microsoft.public.windowsxp.music)
  • Re: Reference User Profile
    ... userprofile includes C:\Documents and Settings (plus there's always a chance ... (no e-mails, please!) ...
    (microsoft.public.access.modulesdaovba)
  • Desktop Apperance
    ... When I right click on my desktop to chance my settings, ... runndll.32xe bad image, then it procedes to tell me to ... Prev by Date: ...
    (microsoft.public.windowsxp.help_and_support)
  • Re: SuSE 9.1 desktop configuration KDE question
    ... Thanks for clarifying my question. ... I've just played around in KDE Control Center (assuming you're using KDE ... Is there a chance you set this on several programs in their ... settings to see if anything helps. ...
    (alt.os.linux.suse)