Re: methods in determining cipher types

From: conrad (conrad_at_lawyer.com)
Date: 01/24/04 

Date: 23 Jan 2004 23:07:17 -0800

Jim Gillogly <jim@acm.org> wrote in message news:<pan.2004.01.24.02.37.03.865491@acm.org>...
> On Fri, 23 Jan 2004 15:07:59 -0800, conrad wrote:
>
> > i'm trying to determine if the type of this cipher
> > is a single columnar transposition cipher?
> >
> > MZDF EZOH VRNT QDET ZKUS BRKG CTHO SKAD LARF JDTP EQUL IBBV QGLM KRUS VZIM
> >
> > But I'm not sure if it might be some deviation from it
> >
> > What's a general measure in checking such things?
>
> A general measure is the Index of Coincidence, which you can look up in
> the FAQ. This one comes out at 0.034 or 0.880, depending on whether you're
> doing the "popular" or "government" version. Either way, it's saying the
> distribution in this cipher is a little more even than you would expect
> from random text, but probably not significantly more random.
>
> If this were ordinary English text the measure would come out at 0.066 or
> 1.73 (depending on the version). A transposition cipher would not affect
> this value -- so either it's not a transposition cipher, or it is but the
> text has been heavily manipulated.
>
> Looking at the frequencies of the characters, we see:
>
> a b c d e f g h i j k l m n o p q r s t u v w x y z
> 2 3 1 4 3 2 2 2 2 1 4 3 3 1 2 1 3 4 3 4 3 3 0 0 0 4
>
> The lack of wxy would be reasonable for a transposition. The high number
> of q's is unusual, but since there are just enough u's to go with them,
> this doesn't rule out the transposition hypothesis. The number of vowels
> is low, and no other letters (like t) have high values. You'd expect 40%
> vowels, more or less, and here we have only 20%. Again, if it's just a
> transposition and the text is not heavily manipulated this is much too low.
>
> There are no significant repeats, so Kasiski doesn't help us. Looking for
> possible periodicity, we see:
>
> 2 0.0333
> 3 0.0404
> 4 0.0429
> 5 0.0273
> 6 0.0519
> 7 0.0391
> 8 0.0342
> 9 0.0529
> 10 0.0200
> 11 0.0333
> 12 0.0833
> 13 0.0744
> 14 0.0262
>
> This suggests that it could be a periodic polyalphabetic with period 6 or 9 or
> 12 or 13.
>
> It could be a digraphic cipher -- the digraphic index of coincidence is not
> impressive, but who knows? There is no XX digraph that would make it inconsistent
> with a Playfair diagnosis.
>
> If you have some reason to believe it started out a complete columnar, then
> either you're stuck with heavily doctored text using q's and few vowels, or
> some additional transformation, like a period 6 Vigenere after the
> transposition was performed. Not, however, normal English, even if a
> simple substitution was done after the transposition.
>

It didn't start out as a columnar
The full cipher text was:

ADEGILNNR AKLOSV
LW52408 5FNW6 17066C SE 227035FOS 313058L K A11301 2GW R421 037O
T A114 183KSF21711 2 IPF8 1 8 036 O KS92807 5RJD7260 2 4KFN02 6147J
VM7 2812 3 G FG8191 14DHO42513 4TGR 5 1 6178AA R318136 OAR621 142G
S X314 26 3JTD42 6224N TU025 2 21YYD1 18251 NF J 924118 GTR21 7213S
JV9 11254 D WC0283 6 7KTR619 275H Y D 72122 4JS F5282 6 9TUF8 152 13K
XX XXXX X XXXXX XXXXX XXXXXXX XXXX X X XXXXXX RA1 2831 3 BXXX XX XXXX
XX XXXTXX X XXXXXXXX XIU EXXJVX KALP X XHBXSY M WXXDRX XCIN XX ZFXG
XXXXXG XXXXXXXXX X XXXPNXXHMXDVSYX XIKXCLETX X JZXBAFOX XWRXU
X XXXXKXXXXXX X X X XXXZRIXXVM XYFW GXXAJ X UOLPXX ECXTN DZXX HXXB
XXX XXEXXXXXXX X XXXXHNUXXTRX O B LXXXDWXYPGJXXV S XIZKFX XMAXC

MZDF EZOH VRNT QDET ZKUS BRKG CTHO SKAD LARF JDTP EQUL IBBV QGLM KRUS VZIM

> So where did this come from?

Friend of mine presented it to me wondering if I could break it
after about a week of reading up on various ciphers
I had only found that the last line resembles a
simple columnar transposition. The rest I can't make any sense of.

Relevant Pages

  • Re: Which paper and pencil cipher to use ?
    ... that it is pointless to design a P&P cipher if you can get access to ... A cipher which requires additional apparatus, ... technique that breaks double transposition fairly fast (in computer ... the book cipher used as page, line, and word indicators, ...
    (sci.crypt)
  • Re: Which paper and pencil cipher to use ?
    ... > that it is pointless to design a P&P cipher if you can get access to ... A cipher which requires additional apparatus, ... A cipher requiring apparatus which cannot be easily disposed of, ... > technique that breaks double transposition fairly fast (in computer ...
    (sci.crypt)
  • Re: Substitution/Transposition - is there a third primitive?
    ... operations in cryptography: Substitution and Transposition. ... I wonder if linear combination (aka Hill's Cipher) wouldn't be ... analysis in ways that simple substitution or transposition does not. ...  One of the common primitives of classical ...
    (sci.crypt)
  • Re: How easy it is to solve "classical" crypto?
    ... >> substitution, fractionation and transposition in one cipher, preferably ... >> without clear fault lines between them. ...
    (sci.crypt)
  • simple symmetric question plz help
    ... from URCPACVTFIGKQGKOON I have to do this using a Caesar substitution ... cipher and a columnar transposition cipher. ... the frequencies and a columnar transposition of 3, ...
    (sci.crypt)
Quantcast