Wednesday, September 19, 2012

Army Command and Administrative Network, IBM Radiotype and APO numbers

In war efficient and secure means of communication are vital for any undertaking. Without the ability to communicate quickly with his underlying units a general will lose control of a battle. Without the ability to control faraway units a military command will see its forces attacked and destroyed piecemeal.

WWII historians tend to focus too much on tanks and aircraft while at the same time neglecting the huge effort made by the warring nations to build up modern communication networks.

During the war the US Signal Corps built up a modern worldwide communications network for US military and diplomatic authorities. This network was called ACAN (Army Command and Administrative Network).

ACAN linked centers in London, Hawaii, New Delhi, Karachi, Chungking, Algiers, Cairo, Basra, Casablanca, Accra, Asmara and other areas with the War Department Signal Center (codename ‘WAR’) in Washington.
 
 
For the rapid transmission of messages several types of equipment were used. Initially hand Morse and Boehme (rapid Morse) equipment, then radiotype and finally the modern radio-teletype (enciphered).

It was thanks to these networks that the Allied commanders could efficiently command their forces all over the globe.

German interception of ACAN networks

These networks were monitored by the German signal intelligence agencies and both plaintext and cipher traffic was intercepted.

Due to the very large number of messages passing on these networks some important information was always available in plaintext transmissions. In addition some of the hand codes used could be ‘broken’.

It also seems that the Germans repeated their ‘Russian Fish ‘success by building a special device that automatically intercepted and printed the teleprinter traffic.

There are few details on this affair but I’ve tried to collate all the available information.

From postwar interrogations it seems that at least 4 agencies intercepted and evaluated this traffic.

1). The Army Ordnance, Development and Testing Group, Signal Branch Group IV C  -  Wa Pruef 7/IV C at an experimental station in Staats.

2). The Signal intelligence Agency of the Supreme Command of the Armed Forces - OKW/Chi.

3). The Luftwaffe Signal Intelligence agency’s LN Abt 350 Referat B in Asnieres, Paris and later Munich-Oberhaching and a special unit operating from Berlin-Schoenfeld.

4). The Army Signal Intelligence agency - OKH/GdNA from bases in Norway and Euskirchen, Germany.

Wa Pruef 7/IV effort:

The Wa Pruef 7/IV C department has been mentioned before in connection with the ‘Russian Fish’ equipment. It was headed by Bau Rat Kierkhoff and according to a postwar interrogation there were about 5-6 engineers, 7-8 other men and 60 odd Nachrichtenhelferinnen (women helpers).

Report FMS P-038 ‘German Radio Intelligence’ says that the Staats station intercepted Russian, British, American and French radio-teletype systems.

 
This traffic was passed on to OKW/Chi and OKH/GdNA for analysis and evaluation.

OKW/Chi effort:

Wilhelm Flicke a member of OKW/Chi mentions in his book ‘War secrets in the Ether’ the interception of US overseas traffic.

In pages 295-6 he says:

Of the "big fry" the USA provided its enemies most amply with information. Among other things a special radio network had been set up in 1942 which covered the entire globe. This was the "WVNA-net" (named from the call-sign of the station in Karachi, India, which was the first one heard.) Most of the exchange of messages could be read currently; it afforded information on American military measures in the Far, Middle, and Near East and in Africa. The following survey shows the extent of the network in November 1942:

WVNA -net

call sign, location, cover name and interpretation 

Washington      agwar = Adjutant General, War Department

                            milid = Attache

                            crypto = Secret communications service

                            victor = Proper name, (head of "crypto") "Signatures: Arnold, Cambela, Groninger, Kroner, Loughry, Marshall, Ohnstaed, Osborn, Reybold, Sommervell, Strong, Ulio.

 

wvna                                    Karachi (India)      speck        =   proper name 

                    Signatures : Jordan (Vice-Consul) ,Wheeler (General and head of the USA military mission).

 

wvmt                           Basra (Iraq)            amsir        = American military section Iraq

                                                                            Signature : Connoly (Vice-Consul)

                                                                        

wvnv                    Cairo  (Egypt)          amsme    = American military Section Middle East

 

wvnt                Asmara (Formerly Italian-East Africa)      amseg    = American military Section

                                                                                                             Signature:Hodges

 

wvmy                Teheran (Iran)                            amrus  = American military Section Russia

                                                                                         Signature:  Ondrick ( Mil.Att.)

 

S9x                          Delhi (India)                      aquila = Cover name for American air forces.

                                                                         ammdel = American military mission Delhi

                                                                        amobsin = American military observer  India

                                                                   Signatures: Tiger, Speck (only there for a short time)

 

nekci                    Chungking (China)     ammisca  =  American military Section Karachi Office in  Chungking         

                                                                               amilat   =   American military attach√©

                                                               Signature : Stilwell ( General and Commander of US Forces in China)

                                                                                         Barret (Military Attache )

                                                                                             Gauss (Ambassador)

 

J7z                                     Kunming (China)           ammkun  =    American military mission Kunming.

 

Bud                          Gura (formerly Italian East Africa)     amgad    = ?

                                                                                               Signature: Bishop,Signals.

Flicke is obviously describing ACAN. This is confirmed by the ‘Signal Corps-The Test’ which also mentions the same stations, for example: ‘After the Signal Corps' first large transmitter in India, at Karachi, had begun operation in April 1942, satellite stations rapidly sprang up.’ and ‘The direct WAR to Cairo (WVNV) circuit, after barely two months of service, was discontinued on Christmas Eve 1942 in favor of relay of the Cairo traffic by way of Asmara (WVNT).’

Unfortunately he doesn’t expand on the kinds of traffic being passed on these links or the codesystems used. However we know that OKW/Chi could decode several US diplomatic codes during the war.

Luftwaffe Chi Stelle effort:

The Luftwaffe Signal Intelligence agency‘s LN Abt 350 Referat B in Asnieres, Paris intercepted USAAF and Naval AF traffic both Morse and radio-teletype. This section reached its peak in terms of importance in 1942-3. Then in 1943 the part that dealt with USAAF ferry traffic was split up and sent to Munich-Oberhaching.

The section of Referat B which was moved to Munich-Oberhaching in 1943 was named Referat B5. There was a large intercept station monitoring USAAF ferry traffic close by.

European Axis Signal Intelligence in World War II’ vol5 gives more details on the operations of Ref B5:

The new Section evaluated all traffic and had the following responsibilities: a) The monitoring of the United States proper, which although it only touched the surface, still furnished an insight into the principal networks of the Army and Naval Air Forces, into training activity, air transport, defense zones, and the activation of new combat aviation units. b) The monitoring of the Atlantic ferry service. The Middle and Central Atlantic routes were monitored by the W/T platoon in Oberhaching and by Luftwaffe Signals Intelligence Service outstations in Spain, which operated under the cover name of ‘Purchasing Agencies’; the North Atlantic route was monitored by the 16th Co., LNR 3 and reports furnished to the Section. c) The monitoring of the American Air Transport Command by the platoon in Oberhaching. d) The monitoring of the RAF Transport Command and of both American and RAF troop carrier commands. The greater part of this interception was also done in Oberhaching.’

US radio-teletype was also intercepted by a LN Abt 350 unit in Berlin-Schoenfeld. The types of traffic intercepted were FF5 and FF6, which refer to 5-unit T/P and Radiotype respectively. A German report from November 1944 says that the Schoenfeld unit had 14 Rhombus installations, an intercept hut with 12 intercept rooms and total strength of 52 men. Intercept equipment was provided by Wa Pruef 7.

In postwar interrogations German personnel mentioned that USAAF T/P ferry traffic was read during the war.

Friedrich, head of Luftwaffe signals intelligence, stated in TICOM I-13:

 KAMERUN’ (CHI Stelle Ob.d.L., Referat B) when located at PARIS-ASNIERES had intercepted and broken U.S. non-morse teleprinter traffic between WASHINGTON and EUROPE. This success was maintained throughout owing to the lamentable insecurity of the operators. The Germans had therefore always known all details of ferry flights, strength of the U.S. Air Force in Europe and a good deal about training and replacements.’ and ‘Yes. It was wireless teleprinter traffic from Washington State Department. No. Not to Europe: to overseas W/T stations. It was a 'Zutraegerverkehr, (traffic passing on signals in bulk). There was an immense volume of traffic. Breaking ceased when a new machine or a new 'Tastschritt’ (keying tempo ?) was introduced. Yes, there was a secrecy device, but it could be broken by the most primitive methods; there were 10 GAF WAAF only to cope with it. The traffic did not carry signals of any tactical importance. No, this was not before, but after, the Americans entered the war. But breaking ceased a long time ago. Yes, a special receiver had to be built. 'As far as I can remember, but I don't want to be bound to the statement, the receiver was built for me by a Feldwebel'. The alphabet was a kind of morse substitution. What should have been A came out as, say, M. No, it was not this impulse business.

The follow-up report I-29 says:

1. Reference was made to Friedrich's previous statements about the breaking of an American "Funkfernschreib" system. See Ticom/I-13. ((Note: as will be seen, the whole difficulty about this claim arose from his use of the quoted term 'teleprinter' when he meant in fact 'undulator', and from a misunderstanding about the terminals of the traffic)). He was asked when the breaking began, and stated that it was shortly after the entry of the U.S. into the war. It was traffic between the War Dept and the regional traffic collecting centers, he could not recall the locations of these centers, but they were all within the United States. They were specifically not in Iceland or Greenland. The transmission was described as teleprinter ((wrong)) short-wave. It was encinhered but was easy to break. It was Morse traffic, not impulse. It was recorded by teleprinter ((i.e. undulator)) on a printed strip as dots and dashes [but see below]. This was enciphered morse, not clear text. He had a special receiver built by a Feldwebel for taking it. it was multi-channel [but see below]; he does not know what kind. The contents of the traffic were: training instructions, development of airfields and details of selective service processes.

2. He was asked whether this was the traffic from which he had said they got details of ferry flights; it was, particularly details of the preparations for flights and of the routes. The traffic dealt with the bringing of aircraft to assembly points, then their transfer to ferrying ports, and finally with their flights to North Africa, Gibraltar, and later straight to Great Britain.

3. Was the morse just scrambled for transmission, or was there a cipher underneath? Didn't know. After unscrambling the channels, did you have clear text, or was there a further encipherment to solve? Reply: there is a misunderstanding. The multichannel scrambling method to which we referred is the method used by the Americans later on, and which the GAF failed to break, because of lack of apparatus. The traffic in question was originally just Plain morse. The new "Tastschritt" affected only the German interception, not us. Asked to define Tastschritt, he said it did not mean "keying speed", but the synchronizing of the intercept equipment with the recorder. He was asked to explain the inconsistency of his referring to the traffic both as teleprinter and as morse. He then drew a picture of an undulator tape, single-channel, and marked off the undulation into successive Morse letters. He said this was what he meant by Funkfernschreib. Each letter was different from the corresponding clear text letter. He could not remember any of the indicators. The preamble gave ample routing instructions, and enabled them to couple the names of towns with call-signs and personal names. The text itself was very stereotyped, especially the addresses. He did not recall whether cleartext was mixed with cipher. He was asked whether they considered it a high-grade system, and replied that they did not, but that there was so much material in it, that if they could have allotted it sufficient time and personnel, they would have got a good deal of valuable information from it. It dealt in addition with the production and development of aircraft. However they had other things to do, and other sources of information, so this materiel was not fully exploited. We asked what these other sources were. He said all of the ferry-flight air-ground traffic was read by III/LN Rgt 3 also in Sicily and gave expected times of arrival and departure, weather, and strength of groups of planes being ferried.

4. It was asked whether the morse signs on the undulator were converted into letters or figures. They were letters. When did the breaking cease? He found this very difficult to recall. It was long before the invasion of France. We asked if it was before the invasion of North Africa. He said it was at just about that time, but he could not remember whether it was shortly before or shortly after. Asked what the change was, he could not recall. Was it to teleprinter or to multichannel. He thought it was not teleprinter as he had never had a receiver built which would take that. Question: Then they were still able to intercept it? Yes. In what form? Doesn't know. Still on tape? He thinks so. Who else intercepted this type of traffic-what other units? OKW did. In Husum? No. Dr. Pickering then said that FNAST 3 (Euskirchen) personnel had told a somewhat similar story (to be published). He replied that it was very likely that this was the same traffic, as KAMERON (unit intercepting for GAF) corresponded with Euskirchen on systems which they both worked on. He agreed with the Euskirchen statement that it was a simple substitution. ((Comment: Not too much credit should be placed in this statement, as he was just adopting a suggestion)).

Voegele, chief cryptanalyst of the Luftwaffe in the West said in I-112: ‘From April to October, 1944, clear radio T/P messages were intercepted regarding a/c movements between America and North Africa. similar messages in cypher with 6 letter indicators were also intercepted but these could not be read.

The Army Airways Communication Service

During the war special support was given to the AACS which belonged to the Airforce. According to the ‘Signal Corps -The Outcome’: ‘AACS differed from ACAN in that it was strictly an AAF organization, manned and operated by the airmen, though Signal Corps men supplied the equipment, engineered and set up the installations, and in the early days of the war often operated the communications lines too, until the AAF could do so with AACS men, who very often merely transferred over from the Signal Corps.’

It seems reasonable to assume that the networks exploited by the Luftwaffe belonged to the AACS. That would explain the traffic dealing with aircraft movements.

Apart from Morse, the AACS was provided with single-channel radio-teletype with automatic enciphering in order to deal with the traffic loads it transmitted daily. The official history ‘Signal Corps -The Outcome’ says: ‘It was the South Atlantic route to Africa and Europe that first got the single-channel RTTY net, along the string of Caribbean islands to the bulge of Brazil, across the South Atlantic via Ascension Island, reaching Dakar by mid-1943.’

OKH/GdNA effort:

The German Army’s Signal Intelligence agency - OKH/GdNA had a special group that intercepted high level enemy radio-teletype traffic. This was Group VI split into Referat 1 operating in the East and Referat 2 in the West.

The Ref 1 unit intercepted Soviet multichannel radio-teletype, both plaintext and enciphered, with special equipment. Members of this unit were captured in May ’45 by the Allies and taken to the UK with their equipment.

The Ref 2 unit was based in Euskirchen, Germany. This was the base of Feste 3 (Feste Nachrichten Aufklärungsstelle/Stationary Intercept Company).

According to report ‘CSDIC 1717’ Group VI Ref 2 was headed by inspector Heller and was divided in Ref 2A which evaluated the British and American T/P and automatic Morse traffic and Ref 2B that intercepted it. Roeder, head of Group VI at the end of the war, stated in TICOM I-99 that the Western unit had about 15 soldiers and 30 helpers plus 10 receiving apparatus.

US traffic was also picked up by Feste 9 in Norway. In report CSDIC/CMF/Y 40 it is stated that Feste 9 intercepted US traffic both manual and automatic from domestic bases and stations abroad (Atlantic area, Caribbean, Middle East, India).

Some of this traffic was enciphered with systems that the Germans had solved. These were the War Department Telegraph Code, the Division Field Code and the M-94 strip cipher.

There was a special device ‘funkfernschreibverkehr’ used by funkmeister Rudolph ‘an expert on WT TP intercept’.  This is a reference to the interception of Baudot radio-teletype and Radiotype.

IBM Radiotype, APO numbers and promotion letters

In order to build a modern communications network the Signal Corps wanted radio-teletype units with automatic enciphering and deciphering capability. In 1942 they did not have such equipment so they had to settle for a similar machine called radiotype.

From the IBM website:

The first working model of the Radiotype was fabricated in 1931 in the laboratory of Radio Industries Corporation under the direction of Walter S. Lemmon, who was then the company's president, Clyde J. Fitch, an engineer, and A. M. Nicolson.

……………………………………………………………………………………………………..

In 1935 Admiral Richard E. Byrd successfully sent a test Radiotype message 11,000 miles from Antarctica to an IBM receiving station in Ridgewood, New Jersey. Six years later, IBM lent Radiotype machines to the U.S. Signal Corps for tests between Washington, D.C., and Dayton, Ohio. These tests were conducted by Albert Holt, an IBM field engineer in the Radiotype Division. With the U.S. entry into World War II, the Signal Corps ordered quantities of the Radiotype machines to equip its stations in San Francisco, Honolulu, Panama, Puerto Rico and elsewhere.’

The ‘Signal Corps-The Test’ says: ‘The International Business Machines Corporation had worked out an imperfect solution involving equipment that the firm called radiotype, using, unfortunately, not the standard five-unit teletypewriter code but a special six-unit code. Like a narrow gauge railroad adjoining a standard line, this special code necessitated much hand labor at conversion points where standard teletypewriter texts had to be shifted onto radiotype circuits, and vice versa. Moreover, the standard automatic cipher machines could not function with the six-unit system. Notwithstanding these inconveniences, the Signal Corps early in the war began making use of radiotype, leased from IBM. It was another step in the right direction, toward automatic, high-speed, heavy-duty communications for the Army.’
 
 
According to the same source the use of radiotype was extensive during the war: ‘Radiotype would continue to be used considerably. Not till September 1943 would the Signal Corps stop its procurement in favor of radioteletype and not until May 1945 would the Army take its last radiotype out of service (on the WAR-Accra circuit). Then the triumph of radioteletype would be complete.

The Germans were able to exploit the internal US traffic from 1941 onwards. The Army intercepted it from Euskirchen and from Norway.

By monitoring the internal US radio traffic the Germans could follow the activation and movement of units through their APO (Army Post Office) number. It was understood that divisions sent to West coast harbors went to the Pacific theatre while those sent to East coast harbors went to the Atlantic theatre. Valuable intelligence was also gained from officer’s promotion letters.
 
 
Initially this traffic was Morse but the introduction of radiotype forced the Germans to build specialized intercept equipment.

FMS P-038 says: ‘In the spring of 1942 a new transmitting technique was introduced in American long-distance communication (both domestic and foreign) that dried up this excellent source of German intelligence. The Euskirchen station, which was charged with cryptanalysis of this traffic, solved the riddle within one week, however, by means of tape recordings and systematic analysis. It was finally discovered that the process used was a rapid system of wireless telegraphy which differed from the usual method by the number of current impulses. This was the ‘Radiotype’ method. A tremendous number of military and business messages were soon intercepted. After a short while the receiving operators were able to ‘read’ the message tapes as fast as Morse code. Fortunately, after a pause of one week, military messages in clear text became more frequent for a time. This mistake was not discovered by the Americans until later, at which time they began to encipher these mechanically transmitted messages. Since it was no longer possible to solve them, work on these messages was discontinued.’
 
 
This may refer to the introduction of the SIGCUM cipher machine in early ’43 and its immediate withdrawal due to a security problem. Or it could refer to operator mistakes described in ‘The Signal Corps-The Outcome’: ‘The on-line features had a serious security disadvantage, however, in that the operator on the transmitting end sometimes forgot-when passing from unclassified traffic sent in the clear to classified-to flip the switch that would connect the on-line crypto equipment.’

The Germans did not always get good intelligence from ACAN. David Kahn says that the Allies managed to deceive them regarding the divisions sent to Britain in 1944 by sending fake radio messages. However he doesn’t provide more details.

On the other hand a member of Feste 3 named Wingender states in TICOM I-76 that even ‘fictitious stations and traffic’ were recognized thanks to violations of radio discipline and cipher security.

Radio-teletype and SIGCUM

Radiotype was only a temporary solution and from 1943 it was being replaced by regular 5-unit (Baudot) radio-teletype.

In order to protect this traffic the Americans developed a cipher attachment that automatically enciphered and deciphered the traffic. This device was called Converter M-228 or SIGCUM and was introduced in January 1943.

Its initial debut was not successful as a flaw in its security was found and a decision was made to delay its entry into service for several months.

It was finally put into use in April 1943. From then on ACAN teleprinter networks would be secure from eavesdroppers.

Unanswered questions

Although the information we have is enough to form a rough understanding of German operations there are many missing elements.

We lack details on the history and performance of the German agencies regarding their interception of ACAN.

What kind of intercept equipment did they use? How much traffic did they intercept? How much of it could they decode?

Did they attack the SIGCUM traffic cryptanalytically or simply use it for traffic analysis?

Another important question is whether the Radiotype system was used with a cipher attachment of some sort. Was that the simple substitution that the Germans talked about? We do know that standard cipher attachments could not be used on it because of its 6-unit operation.

Building IBM: Shaping an Industry and Its Technology’ says in page 348: ‘The use of Radiotype coupled directly to encryption and decryption equipment, is reported by J. C. McPherson. 8 August 1991: discussion with E. W. Pugh.’

Let’s hope that some of these questions will be answered in the future.

Sources: The Signals Corps trilogy (US Army publications), FMS P-038 ‘German Radio Intelligence’, ‘War secrets in the Ether’, CSDIC 1717, CSDIC/CMF/Y 40 , TICOM reports D-4, I-13, I-29, I-42, I-64, I-65,  I-76, I-78, I-99, I-104, I-109, I-111, I-112, I-149, Cryptologia article: ‘The Sigcum story: cryptographic failure, cryptologic success’ , IBM website, ‘Hitler’s Spies’, ‘European Axis Signal Intelligence in World War II’ vol4 and vol5

4 comments:

  1. let's also recall IBM working with the Nazis, and providing the backbone of THEIR communication system!

    ReplyDelete
    Replies
    1. I think that the German communication equipment was all domestically produced. If you are referring to IBM punch card machines that's another issue.

      Delete
  2. Allied vs. German Cryptology David Kahn Video 57 minutes From Oct. 26, 1995
    Didn't have many "views" on the cspan site.

    http://www.c-spanvideo.org/program/67946-1

    Professor Kahn compared Allied and German cryptography, encoding, and cryptology, decoding. He examined why German coding activity was inferior to Allied coding activity. (later in video)Professor Burke spoke about how the needs of U.S. cryptographers led to the development of machines which were the precursors of the modern computer.

    -kurt

    ReplyDelete
    Replies
    1. 'He examined why German coding activity was inferior to Allied coding activity. '

      Kahn's books were written a long time ago with the limited information available. New information has superseded his research.

      Delete