Numbers station - E10 Echo Zulu India - full transmission

Quite a noisy recording, but all the elements are there. E10, using callsign 'Echo Zulu India' sending a 20 group message. Broadcast near 6840 khz 1700 UTC 1/3/2008.

E10 is believed to be operated by Mossad.

The callsign ends 2:27 into the recording, then goes into the preamble and the message is read and repeated.

The message is transcribed as follows. The chances of a crack are minuscule, but anyone is free to try:

BBEDI ZDSZA NREIG ZMSQF NNYSM
PLNXD FZGDY EEWIN UOORA FGDPD
LDCGW INWOL QSOHJ QDYGE MJBEX
EHGYK OZBGV TYOJM BKDXI MLWOB

(There may be one or two errors as the recording is unclear)

Note: there is a slight difference in intonation between letters in the start/middle of a group, and the last letter of the group (which has a falling intonation). This is audible when a group has the same letter in two different positions, including one as the last letter (e.g. the 10th group: 'FGDPD').

This technical effect makes the automated voice sound more natural. Previously, it was thought that only E3 (The Lincolnshire Poacher) used a system with changing intonation, although it seems that E10 uses such a system too. This is a technical gloss which indicates an agency with good infrastructure, using similar technology to civilian automated telephone exchanges.

For those who have stumbled across this page and have no idea what is going on, please see http://en.wikipedia.org/wiki/Numbers_station

UPDATE: Very heavy EZI traffic on 7/3/08. Day after gun attack in Jerusalem, so that may explain increased Mossad activity. Message sent near 2130 UTC had numerous groups - continued broadcasting until 2156 UTC.

Category:

News & Politics

Tags:

• numbers
• station
• radio
• shortwave
• espionage
• spy
• code
• cryptography
• intelligence
• E10
• Mossad
• Israel
• Echo
• Zulu
• India

[LCduT]

Fine Dieselscience. This is academic anyway, as your comments aren't about a true one time pad (which requires a random key). As in my first comment, once the real OTP is too difficult to handle people use lesser systems. A running key cipher added to another cipher is still not an OTP and is still breakable. Even if the running key text is used once.

As for Ana Montes, I believe she used a computer program with keys handed physically by disk. Why would they need The Grapes of Wrath too?

[Erectoralporicy]

I don't know why people are talking about deciphering this. It litterally is impossible. Books are used once, the instructions are believed to be pages you rip out. What you are left with is a transcribed messsage. Unless you have a book, you're not figuring this out.

[onelasssttime]

Wow, Ty for your lucent comments -- LCduT

[dieselscience]

Keep researching, books are great when used only ONE time. BUT the assumption "legible result" should not be made IE they're NEVER plain text, they're brevity words and code words! I can't elaborate and still be 'legally non-disclosing.' Analysis with computers would work quickly with skip codes but not all codes are skip cipher. Some characters within a set & sets within a message are filler. Remember this, different languages have unique letter combination patterns-not so with numbers.

[LCduT]

...cont'd. Even before supercomputers you could run common phrases (e.g. 'and the', 'and a', 'was not' etc) through a ciphertext. Any fragments of legible text (even if hidden by simple substitution ciphers) in any language would be noted. Once enough of these common phrases produced legible results from the ciphertext, trial and error (guessing the ciphertext and seeing if the running book key produced was legible) would result in a crack. That's the weakness of not using a random OTP key.

[LCduT]

If the Ana Montes spies used book text as an OTP key, that was very foolish. It's not just Project Gutenburg that digitises books. Let's assume a sigint agency has every available book digitised. All they would have to do is run the ciphertext through every possible start position in every possible book. An easy task for a computer. Any result (possibly only one) that isn't gibberish will probably be the decoded plaintext.

[dieselscience]

YES! You need something to read a micro dot, any kind of lens device can be found in a search. You're a 'spy leaving the home country' Your manager gives you a bible and says "23rd of this month use page 231" I send a numbers transmission on the 23rd, it contains the message. We talked about the details before you left(odd number=vowel,etc.) That page will never be used again. Look up Anna Montes case, I think they used Grapes of Wrath MULTIPLE TIMES! 1 time pad=ONLY 1, pattern=cracked=busted!

[LCduT]

A page of the Bible or any known book as a pad? I hope you don't mean the actual text of such a book (as opposed to a random key hidden in, say, a microdot), as that would be a running key cipher, which almost any modern supercomputer could crack very quickly.

[dieselscience]

When I was in the *US* Army...what the C.I. crews do is look/listen for patterns. Cracking, as you say, 'technically' possible but it would take literally years without minimal knowledge of the (number) set. We know the patters of many, just NOT the pad (or S.O.I.) AND DEciphered messages don't have to be legible to those they're not intended for.

Furnish the pads? EASY, any given page of a Bible, Koran or phone book - never get caught with something that appears like a pad, its how you use it.

[LCduT]

It's mathematically impossible but the catch is the system has to be handled perfectly.

Possible weak links in handling:

- The difficulty in producing truly random keys. If the key is pseudo-random (or worse, a running-key cipher of legible words) a crack is technically possible

- The practicalities of furnishing an agent with enough pad. This requires dangerous meetings or - worse still - some modern pads are reportedly digital which opens a massive interception risk