|
Home Page Kryha Home Page |
IntroductionThe German networks in Argentina during World War II were organized into three networks, codenamed RED, BLUE, and GREEEN. RED was the operational network of the SD (Sicherheitsdienst), or Security Service. It focused on the political aspect and communicated with the RSHA (ReichsSicherHeitshAuptmt or Reich Security Intelligence Service) in Berlin. The RSHA controlled the Gestapo and the SD, and later even the Abwehr (German military intelligence). In 1943 (from January to December), the RED network was encrypted with the Kryha Liliput. When the machines became unusable, RED switched to Enigma. Initially, it was encrypted using double transposition. The US Coast Guard and RED trafficThe RED traffic was analyzed and decrypted by the U.S. Coast Guard's cryptographic service, then headed by Elizabeth Friedman (W.F. Friedman's wife). This service, originally independent, had been part of the Navy's Op-20-G since late 1941. As previously reported, the RED network used a double transposition system that had also been broken by the Coast Guard. Message 548 from January mentions the upcoming use of the Kryha Liliput to replace the current system. In message 584 (also encrypted using double transposition) sent from Argentina, Elizabeth's team was able to read: "...Following messages all enciphered with Liliput. Because adjustment number wheel forgotten, stop all even numbers. Starting position number 1 and A opposite A…". Thus, American cryptanalysts learned that future messages had to be encrypted with the Kryha called Lily [Liliput?]. They possessed a copy (a Kryha standard but interoperable model). They also knew the wheel's configuration and the initial sector. However, they did not know the alphabets used. US Coast Guard broke Kyha trafficIntroductionThe eight messages sent in early March yielded 2,300 letters, which allowed the message to be deciphered after reconstructing the alphabets; the wheel configuration had been preserved. The method used is broadly that of Friedman, with the added difficulty that the encrypted data consisted not of a single cryptogram but of eight messages (but in-depth)! Friedman's method requires dividing the cipher letters into different groups, each using the same alphabet. To achieve this, the Elizebeth's team encrypted the letter A 626 times (626 = 26 x 26 = key size) using the same wheel configuration they had found, but with direct but reversed alphabets (ZYX...A). This reference cipher allowed them, to distribute the letters into the 26 alphabets.
Encrypt the letter A 626 times
$ cat marks3.sh
python3 -c 'for i in range(676):print("A",end="")' | \
python3 kryha_tui.py \
-i ZYXWVUTSRQPONMLKJIHGFEDCBA \
-e ZYXWVUTSRQPONMLKJIHGFEDCBA \
-s 1,3,5,7,9,11,13,15,17,19,21,23,25,27,29,31\
,33,35,37,39,41,43,45,47,49,51 -m -c
$ sh marks3.sh | fold -w 26 | sed 's/[A-Z]/& /g'
H O U B J P V D J Q X D K R Z F M S Z G N V B I O X
E L R Y G M S A G N U A H O W C J P W D K S Y F L U
B I O V D J P X D K R X E L T Z G M T A H P V C I R
Y F L S A G M U A H O U B I Q W D J Q X E M S Z F O
V C I P X D J R X E L R Y F N T A G N U B J P W C L
S Z F M U A G O U B I O V C K Q X D K R Y G M T Z I
P W C J R X D L R Y F L S Z H N U A H O V D J Q W F
M T Z G O U A I O V C I P W E K R X E L S A G N T C
J Q W D L R X F L S Z F M T B H O U B I P X D K Q Z
G N T A I O U C I P W C J Q Y E L R Y F M U A H N W
D K Q X F L R Z F M T Z G N V B I O V C J R X E K T
A H N U C I O W C J Q W D K S Y F L S Z G O U B H Q
X E K R Z F L T Z G N T A H P V C I P W D L R Y E N
U B H O W C I Q W D K Q X E M S Z F M T A I O V B K
R Y E L T Z F N T A H N U B J P W C J Q X F L S Y H
O V B I Q W C K Q X E K R Y G M T Z G N U C I P V E
L S Y F N T Z H N U B H O V D J Q W D K R Z F M S B
I P V C K Q W E K R Y E L S A G N T A H O W C J P Y
F M S Z H N T B H O V B I P X D K Q X E L T Z G M V
C J P W E K Q Y E L S Y F M U A H N U B I Q W D J S
Z G M T B H N V B I P V C J R X E K R Y F N T A G P
W D J Q Y E K S Y F M S Z G O U B H O V C K Q X D M
T A G N V B H P V C J P W D L R Y E L S Z H N U A J
Q X D K S Y E M S Z G M T A I O V B I P W E K R X G
N U A H P V B J P W D J Q X F L S Y F M T B H O U D
K R X E M S Y G M T A G N U C I P V C J Q Y E L R A
If we inspect this pseudo-cryptogram, since we have only encrypted As, it follows that two identical letters have been encrypted with the same alphabets. For example, the 5th and 9th letters are Js, so these positions use the same alphabet. Similarly, we notice that each column uses all 26 alphabets. We can create a set of punched cards as follows. For each letter of the eight real cryptograms, we create a punched card containing the encrypted letter and the letter of the pseudo-cryptogram AT THE SAME POSITION. We can then classify all the letters indexed by the same letter of the pseudo-cryptogram, which are therefore part of the same alphabet. We obtain approximately 90 letters per alphabet. Then we apply the rest of Friedman's method. Note: it should be noted that the eight messages were in-depth. The key of the messages (including No. 586)Elizabeth's team discovered that, contrary to the Kryha manual, the operator advanced the wheel before encrypting. Here is the complete key and the decryption of message 586.
The message 586The Cryptogram: $ cat MSGS/marks.cry QGSSO KQHBX QYEAF DLGVG DPJSL S KWXDT FMJVV XJQBB JSTEX FSLBN T OOOXL CTXLU RSJZP CVEZW DZSQJ Z ABGJF LTRCC KREVV VJSIT QLJAT X ZKKUU PEAWV PYQZR CUKFZ PMUPB S FADNP RDZNO GVKYU PDMQL TMDTG A IINOA VREDL BOIVN HLQED RIULT M YNVGL DTVMC VIO The plain text (raw):
$ cat marks.sh
#!/bin/sh
# marks.sh
python3 kryha_tui.py \
-i SVGDWMQCJAHTEXRPNFIKUBZLYO \
-e ESPFITVONALDCHRBGJKMQUWXYZ \
-s 1,3,5,7,9,11,13,15,17,19,21,
23,25,27,29,31,33,35,37,39,41,43,45,47,49,51 \
-a AA \
-f 1 \
-m \
-d
$ cat MSGS/marks.cry | sh marks.sh |python3 groupe.py
HGQSN NSHIF FSWEG NAQEN GLAND XXBIS JFXNO RONHA JLAEN GSKUE
STEOH NEGEL EITXC ONVOY BILDU NGHOE HEJFX NORON HAJXD ANNLA
ENGSJ USAJK UESTE BISSE QZIGG RADXD ANNPA RALLE LBREI TENGR
ADBIS JORKN EYJXQ UELLE XXJVM JAUFH OLLAE NDXDA MPFER
The interpretation of the message 5/36 nn shiffsweg naq england xx bis j f. noronha j laengs kueste ohne geleit x convoy bildung hoehe j . noronha j x dann laengs j usa j kueste bis seqzig grad x dann paralll breitengrad bis j orkney j x quelle xx j vm j auf hollsend x dampferTranslation : From : Argentine To : Berlin February 29, 1943 5/36 Shipping route to England : to F. Noronja along the coast without escort. Convoy assembly in latitude of F. Noronha then along USA cost as far as 60° then parallel to latitude, parallel as far as Orkneys. Source "V" man on Dustch steamer. ReferenceArticles & Books
Web sites
|