![[JoGu]](../../JGU.gif) |
CryptologyA Short Ciphertext |
|
|
CryptologyA Short Ciphertext |
|
MDJJL DSKQB GYMZC YKBYT ZVRYU PJTZN WPZXS KCHFG EFYFS ENVFW KORMX ZQGYT KEDIQ WRVPM OYMQV DQWDN UBQQM XEQCA CXYLP VUOSG EJYDS PYYNA XOREC YJAFA MFCOF DQKTA CBAHW FYJUI LXBYA DTT
The KASISKI test finds no reptitions of length 3 or more. It finds 16 repetitions of length 2 and no eye-catching pattern. The common factors 10 or 20 could be a hint at the correct period, but repetitions of length 2 are not overly convincing.
| Repetition: | DS | SK | GY | YM | CY | BY | YT | TZ |
|---|---|---|---|---|---|---|---|---|
| Distance: | 98 | 28 | 47 | 60 | 100 | 125 | 40 | 8 |
| Repetition: | GE | FY | OR | MX | QW | DQ | AC | YJ |
| Distance: | 60 | 94 | 60 | 31 | 12 | 50 | 40 | 21 |
The inner coincidence index of the text is 0.0386 and doesn't distinguish the ciphertext from random text. The first 40 values of the autocoincidence spectrum are
| κ1 0.0270 | κ2 0.0203 |
κ3 0.0541 | κ4 0.0405 |
κ5 0.0405 | κ6 0.0338 |
κ7 0.0405 | κ8 0.0676 |
| κ9 0.0270 | κ10 0.0473 |
κ11 0.0270 | κ12 0.0676 |
κ13 0.0405 | κ14 0.0473 |
κ15 0.0541 | κ16 0.0541 |
| κ17 0.0203 | κ18 0.0203 |
κ19 0.0608 | κ20 0.0473 |
κ21 0.0473 | κ22 0.0135 |
κ23 0.0541 | κ24 0.0270 |
| κ25 0.0338 | κ26 0.0405 |
κ27 0.0541 | κ28 0.0811 |
κ29 0.0338 | κ30 0.0338 |
κ31 0.0405 | κ32 0.0203 |
| κ33 0.0068 | κ34 0.0473 |
κ35 0.0473 | κ36 0.0270 |
κ37 0.0405 | κ38 0.0066 |
κ39 0.0203 | κ40 0.0473 |
Values above 0.06 occur for shifts of 8, 12, 19, 28, the latter being the largest one. This makes a diffuse picture, giving slight evidence for a period of 28.
Finally let's try SINKOV's test. It gives as its first 40 values:
| φ1 0.0386 | φ2 0.0413 |
φ3 0.0386 | φ4 0.0492 |
φ5 0.0421 | φ6 0.0441 |
φ7 0.0433 | φ8 0.0471 |
| φ9 0.0330 | φ10 0.0505 |
φ11 0.0265 | φ12 0.0591 |
φ13 0.0333 | φ14 0.0486 |
φ15 0.0444 | φ16 0.0410 |
| φ17 0.0280 | φ18 0.0395 |
φ19 0.0439 | φ20 0.0589 |
φ21 0.0357 | φ22 0.0264 |
φ23 0.0476 | φ24 0.0548 |
| φ25 0.0507 | φ26 0.0359 |
φ27 0.0444 | φ28 0.0488 |
φ29 0.0368 | φ30 0.0622 |
φ31 0.0312 | φ32 0.0323 |
| φ33 0.0091 | φ34 0.0294 |
φ35 0.0429 | φ36 0.0611 |
φ37 0.0541 | φ38 00.0307 |
φ39 0.0256 | κ40 0.0542 |
The values for 12, 20, 30, and 36 stand somewhat out, followed by the values for 24, 37, and 40, then 10 and 25—again there is no clear favorite.
Let's discuss the candidate values for the period.
| Period? | Pros and cons |
| 8 | φ(c) should be slightly larger (weak). Only 3 repetition distances are multiples of 8 (weak). κ8 and κ16 are good, κ40 is weak, κ24 and κ32 are prohibitive. φ8 is weak, φ16 and φ32 are prohibitive, φ24 and φ40 are good. |
| 10 | φ(c) should be slightly larger (weak). 7 repetition distances are multiples of 10 (good). κ10, κ20, and κ40 are weak, κ30 is prohibitive. φ10, φ20, φ30, and φ40 are good. |
| 12 | φ(c) should be slightly larger (weak). 4 repetition distances are multiples of 12 (good). κ12 is good, κ24 and κ36 are prohibitive. φ12, φ24, and φ36 are good. |
| 19 | 0 repetition distances are multiples of 19 (prohibitive). κ19 is good, κ38 is prohibitive. φ19 and φ38 are prohibitive. |
| 20 | 6 repetition distances are multiples of 20 (good). κ20 and κ40 are weak. φ20 and φ40 are good. |
| 24 | 0 repetition distances are multiples of 24 (prohibitive). κ24 is prohibitive. φ24 is good. |
| 28 | Only 1 repetition distance is a multiple of 28 (weak). κ28 is good. φ28 is weak. |
| 30 | 3 repetition distances are a multiples of 30 (good). κ30 is prohibitive. φ30 is good. |
| 36 | 0 repetition distances are multiples of 36 (prohibitive). κ36 is prohibitive. φ36 is good. |
| 37 | 0 repetition distances are multiples of 37 (prohibitive). κ37 is prohibitive. φ37 is good. |
To assess these findings let us invent an ad-hoc procedure and weigh the values »good« as +1, »weak« as 0, and »prohibitive« as -1. Note that 3 repetitions for period 8 are weaker than 3 repetitions for period 30. The candidates 19, 24, 36, and 37 have negative weights, the candidates 8 and 28, zero weights. We skip them in the first round. Positive weights have 10 (3 of 9), 12 (3 of 8), 20 (3 of 5), and 30 (1 of 3). We rank them by their relative weights: 20 with score 0.6 = 3/5, then 12 with score 0.375, then 10 and 30 with scores 0.333.
The most promising approach to further cryptanalysis starts from the hypothetical period 20.