Secrets of Short Wave radio espionage by Don Schimmel:


Suppose you were going to set up a clandestine system of communications with your agents in the field. It seems logical that the overall traffic would be composed of three main categories: First, there would certainly be some practice traffic. This would be required to accustom the trainees to listening for and tuning-in their respective broadcasts so as to receive traffic addressed to him/her. This practice traffic would provide valuable experience and improve their proficiency in copying messages as well as in decryption or decoding, as the case may be.

Now a second type of traffic could very well be deception traffic. What better way to confuse opposition traffic analysis and cryptanalysis efforts than by throwing in traffic that was nonsense? Dummy messages could be made up with normal headings but with texts of purely random digits as just one of the ways of generating non-valid traffic. The absence of a pre-specified indicator group would tip off the agent that the traffic was bogus and he need not finish copying it. Eventually the overall plan should probably include a phasing out of the practice messages after a suitable breaking-in period of the agent and replace those messages with dummy traffic and then blending in the live traffic.

The third type of traffic would of course be valid messages. I doubt that every message would be of the "BLOW UP THE DAM" class. Instead there would probably be the myriad of details to be accomplished such as selection of secure meeting places, instructions regarding financial arrangements, notification of mailing addresses, crypto matters, directions concerning meeting schedules and means of recognition/identification of the agent handler. Any of the thousand and one real-life concerns of a spy organisation and its clandestine personnel in place overseas. This getting settled-in may require many months, and in some cases, perhaps years, before all of the necessary preliminary assignments have been carried out.

I would suspect that only then would the agent be activated to provide reporting on a regular basis. Bear in mind that it stands to reason that all the time he is carrying on his clandestine activities, it would seem imperative he must also maintain a normal profile with some type of work plus engaging in other appropriate daily activities that "fit in" thus assisting in building and maintaining his cover.

A possible reason for so many one-way voice broadcasts is that perhaps it's just simply not a good idea from a security stand-point for these agents to possess transmitting equipment. The solution, therefore, is to provide a common type radio, one with SW bands, which would certainly not arouse any suspicion when viewed by agent acquaintances. When not permitted to use radio for agent to headquarters traffic, a possible reply channel could be via mail where an innocent appearing text serves as a means of conveying secret writing or a message in an open code system.

Lastly, let's consider the cryptographic system to be employed. It goes without saying that not only do we want a good secure system, but one which is also not too complex to use. In going through the multitude of press accounts regarding arrested spies there are so many references to the use of the one-time pad system. Doesn't it make sense that any spy outfit worth its salt is not going to use some rinky-dinky system? During past years I have seen frequent boo-boos occurring during the transmissions of "Spy Numbers" messages and I want to mention several that were particularly amusing or unusual.

 While searching the 13 MHz band one evening I came upon a very weird sounding voice transmission on 13640.6 kHz. After listening to it for several minutes, I finally realised it was a Spanish female announcing 5-figure groups, but the reason it sounded so strange was because the tape was being played backwards. The entire message was transmitted this way and then repeated the same way. Apparently the operational personnel were not monitoring their own signal and consequently they did not catch the error.

Another messed up transmission took place on 13390 kHz with cut number (5-character groups) traffic being sent in CW which looked like the format of Cuban Foreign Ministry traffic. Right underneath that station another CW carrier was feeding through with another cut number (4-character groups) message. On 13429 kHz, a Spanish female was sending 5-figure groups, but the audio was very distorted. The signal abruptly shifted up approximately 20 kHz, and then shifted back down to 13429, back up to 13448 again, and back down to 13428, and ultimately back up to 13448 again. The voice transmission terminated but it sounded as if the loose end of a tape was flopping. This finally stopped and they gave up for that night. There was a high hum level on the carrier which seems to be a very common characteristic of broadcasts suspected of coming out of Cuba.

Then there was the time a Spanish speaking female was transmitting on AM with a 5-figure groups message and shortly after the "Final Final" phrase the carrier came back on the air carrying the musical theme introduction for Radio Havana. After a few minutes of this music, a Spanish male identified the station as "Radio Havana Cuba" and a Shortwave Broadcast program commenced. This program was only broadcast for a very few moments when the carrier was abruptly taken off the air. I have already described one of the double-header intercepts I made but I also had a triple-header. I had tuned in a CW station on 16446.8 kHz which was sending a callup indicating a message was forthcoming.

After sending the text twice, another message was sent and it was also repeated. During the repeat of the second message, I became aware of another CW transmission underneath the first, and it was also in 5-figure groups. In tuning off to the side of the second signal by just a hair, I detected yet another signal under the other two, this one was a voice transmission with a Spanish female giving a 5-figure groups message.
The CW carriers went off the air as each completed the traffic and finally the voice signal went off the air. I wonder if this triple transmission was the result of some kind of coupling between control lines carrying the signals from their source to the transmitter building?

The first type to be looked at is the SS/YL 4F broadcast. Although the mode was initially thought to be AM, many of the 4F schedules are actually reduced carrier USB, broadcast on 2 frequencies simultaneously. The callup includes a trinome (probably an addressee indicator) which is repeated 3
times followed by digits 1-0 and the entire sequence is repeated over and over for about 8-10 minutes. Ten tones are then sent followed by the group count (usually repeated twice) and then into the text. At the end of the message the group count is given again and the message is repeated. At the completion of the repeat "FIN" is sent and the station goes off the air. Sometimes the "FIN" does not appear.
EXAMPLE: 662 662 662 1234567890
GRUPO 56 GRUPO 56
(TEXT IN 4F GROUPS)
REPITO GRUPO 56
(TEXT REPEATED)
FIN

Some of these transmissions are reportedly coming from a Virginia location, a US Army installation which also is supposed to house the transmitters for the US State Department. The airwaves are full of this next type which has both SS/OM and SS/YL announcers sending 5F groups. The normal callup consists of the Spanish word for Attention followed by a trinome (probably an addressee indicator) and then the group count. This sequence is repeated for several minutes and then into the text. At the end of
the message, FINAL FINAL is sent indicating the message has been completed . The message is usually repeated on subsequent schedule on a different frequency.
EXAMPLE: ATENCION 12356 (TEXT IN 5F GROUPS) FINAL FINAL
The 5F transmissions &e reportedly from a location near Havana, Cuba but there is also some evidence that certain of these broadcasts are from other locations.

I would like to include some excerpts from "The FBI-KGB War, A Special Agent's Story," by Robert J. Lamphere and Tom Schachtman. In their book, the authors describe a numbers transmission called the "Allo" broadcasts. "The 'Allo' broadcasts were known to originate in Russia on different frequencies and at various times during the week. An announcer would come on the air and say, in accented English, 'Allo, Allo,' and then recite groups of five-digit numbers. These broadcasts had been monitored by the NSA and other communications intelligence people for some time; the number-groups had never been deciphered." Perhaps, many of you will recall the stories in the press, in 1957, of the arrest of Soviet agent, Rudolf Abel, and the subsequent information relating to the discovery by a newsboy of a hollowed-out nickel he had received in change from a customer. The nickel hid a piece of microfilm on which was an enciphered message of 207 five-figure coded groups.