The first table is the easiest one. The alphabet is numbered A = 01 trough Z = 26 and 00 for a space. L/F 88 (letters/figures) is used just before and after figures and these figures are converted into themselves, written three times to exclude errors (assigning two-digit values to numbers is not advicable as these are prone to single-digit errors). CODE 99 is an optional codebook prefix, followed by a codebook prefix for fixed length code. The second row, from 30 onwards, is optional. If you don't use any optional characters, you can use an "X" as a decimal point and a "Y" as a comma within figures mode.
This approach has the advantage of easy use in the field and doesn't require the help of a complex or hard to remember conversion table. Of course, this method can be expanded with other characters, special tokens, upper and lower case letters, or adapted for any other language. A downside is that a message will be longer due to the two-digit values for all letters. From the security point of view there is no difference with the other conversion systems, described further on, if correctly encrypted with one-time pad. This method is very suitable for small messages and unexperienced personnel. An example conversion:
A second method is to use a straddling checkerboard. The most frequently used letters are converted into one-digit values. All other letters, the space and letters/figures have two-digit values. This reduces the size of the message considerably (+/- 150% of the plaintext against >200% for the previous method). In a straddling checkerboard, the first row contains the single digit characters and is numbered 0 to 7. In the example below we use the eight most frequent English letters SENORITA. The second and third rows contain the remaining letters, L/F 98 and SPACE 99. L/F is used just before and after figures and the figures are converted into themselves, written three times to exclude errors. Use an "X" as a decimal point and a "Y" as a comma within figures mode. Other characters can be replaced by rare combinations, for instance, XX for full stop, ZZ for a slant, YY as optional codebook prefix. The original checkerboard works with columns and rows. If a letter is found in the first row, we take the digit of the column. If the letters is found in the second or third row we take the digit of that row and the digit of that column ( R = 4, H = 85, P = 90 etc).
Presented in a practical table:
An example conversion:
Note that even this very small text is already seven digits smaller than the previous simple method! When decrypting a message, how do we know if we have to read a single-digit or double-digit character? You simply look at the next digit to read. If it starts with 0 through 7, you have a single-digit value. If it starts with 8 or 9, you have a two-digit value and there's one more digit to read for this particular character. The order of the characters can be changed to any desired combination. Some other anagrams for the letters "SENORITA" are "A-NOTE-SIR", "NATO-RISE-", "RAT-NOISE-", or "NO-TEA-SIR". Of course, the letters can be replaced by the most frequent letters of another language. In such case, the second and third rows are again filled with the remaining characters. The position of the two blank fields in the top row can also be changed. For instance, with "NO-TEA-SIR" you will have a blank at 3 and 7. In such case, the second row will carry the numbers 30 trough 39 and the third row 70 through 79. Again, which digits are used for a given character is unimportant and has absolutely no effect on the security of the message, as one-time pad encryption is applied. Many variations on this checkerboard are possible, as long as both sender and receiver agree on a common system. The CT-37 table is an extended version of the straddling checkerboard and includes additional characters. The table is easy to remember. It uses the 7 most frequent English letters "ESTONIA" in the top row. The two following rows are the remaining letters, completed with the "FIG" field. The fourth row contains the "SPACE" and "CODE" field with the punctuation marks between them (less critical to remember). Using the CT37 table is easy. All characters are encoded into their one-digit or two-digit value. To encode figures, always use "FIG" (89) just before and after figures. Each digit is written out three times to exclude errors. You can use spaces and punctuations within the "FIG" mode, for example, "1.5 KG" = "89 111 91 555 89 77 74 ". The "CODE" field is the optional codebook prefix.
In this example we use the code 1234 at the end (which could, for instance, stand for "Send confirmation as soon as possible" in our codebook).
The CT-37-words table uses a mix of letters, words and codes. CODE 6 is a shortened prefix for fixed length codes. The commonly used words "acknowledge", "request", "message", "rendez-vous point", "grid" (coordinates), "send" and "supply" are respresented by a small two-digit code. Omitting one more top-row letter or the CODE field will give another full row which could held 10 more words, expressions or short sentences (CODE could be added to this new row). This approach can reduce the message length enormously if the set of two-digit expressions is selected carfully.
In the CT-46 conversion table, we can use four full rows since we have four unused digits in the top row. This conversion table is not that hard to memorize. In the first row the 6 most frequent letters AEINOR get the single digits. Each next row starts with the remaining digits 7, 8, 9 or 0. The second and third row are the remaining letters of the alphabet. The fourth row contains the SPC and CODE with the signs in between (less important to memorize). The fifth row are simply the numbers preceded by a 0.
The CT-55 table has even more additional characters. L/F 89 is used to switch from Letters (yellow) to Figures (green) and from Figures to Letters. This enables more characters with the same conversion value. The Red fields can be used in both Letters and Figures mode, thus a space, period, etc in a text doesn't require switching to Figures. Numbers are represented by its double code to exclude digit errors. An example: F-16B is converted to 73 89 84 11 66 89 70. CODE is again used as a prefix for four-digit codes and RPT is used to repeat important pieces of text. This conversion table has the 7 most frequent letters, optimized for English, and is suitable for text with more figures and signs.
Of course, there are many other ways to convert characters into numbers. The system of converting characters into digits doesn't need to be secure because the one-time pad provides the security. Any method to convert into digits is good, as long as it is practical, doesn't make the ciphertext too long and isn't prone to critical errors. More conversion tables and manual encryption methods, used by Intelligence agencies, are found on the SAS und Chiffrierdienst website
|
