Wednesday, November 27, 2013

More Seabourne reports available

Randy Rezabek of TICOM archive has uploaded more Seabourne reports. The new ones cover cryptanalysis in the German AF, the OKW Radio Defence Corps and the Signal intelligence Service of the Luftwaffe.

Monday, November 25, 2013

The German intercept station in Sofia, Bulgaria

The German High Command’s deciphering department – OKW/Chi intercepted radio traffic from various stations both in Germany and abroad.

Stations in neutral countries operated covertly, so as not to attract the attention of the Allies.
One such station was based in Sofia, Bulgaria. During WWII Bulgaria followed a pro-Axis policy and declared war on Britain and the United States but did not take part in the fighting.

According to Wilhelm Flicke, who worked for OKW/Chi, an intercept station was set up in Sofia, Bulgaria in January 1940. The station was housed in the former residence of the Communist official Stoitscheff who had fled the country.
Officially it was designated ‘Seismographic and weather reporting station’ but the local authorities knew its true function and cooperated with the Germans. The cover name of the station was ‘Bohrer’, it had about 25-30 men and head of the station was 1st lieutenant Grotz. Emphasis was given on the interception of radio traffic from Turkey and Malta, as well as stations from Egypt, Sweden, Switzerland and the US Armed forces in the Mediterranean.
The station had a direct teleprinter connection with OKW/Chi and in addition there was a courier plane between Sofia and Berlin.

Even though Bulgarian officials helped in setting up the station this does not mean that the Germans held back from attacking their codes. According to Flicke copies of the Bulgarian codebooks were acquired by the Abwehr (military intelligence) station in Sofia.
As the German position in the Balkans began to unravel in 1944 the Sofia station was closed down. This operation did not run smoothly. The equipment was loaded into two freight cars and the personnel sold their unwanted items. With the money earned they bought 80.000 cigarettes that they expected would be valuable back home. However this ‘treasure’ was lost when the railway car was attacked by partisans and the ammunition stored together with the cigarettes burned up.

Moral of the story, never store tobacco and ammunition together, especially if you’re travelling through the Balkans!


I have uploaded TICOM report DF-116-K ‘The German intercept station in Sofia’ - 1948, written by Wilhelm Flicke.

Available from my Scribd and Google docs accounts.

Tuesday, November 19, 2013

British report on German armor piercing projectiles

The very interesting report ADM 213/951 ‘German steel armour piercing projectiles and theory of penetration’ is available from World of Tanks forum user Daigensui.

From page 19 onwards there is a review of the German method of staging and conducting tank round penetration trials. Source of the information was
‘The writer was fortunate in tracing Oberbaurat HENNING TELTZ of Wa Pruef 1 (1X). This man was in charge of the firing of all trials of A.P. Shell against armour plate, masonry, concrete and soil and was responsible to Oberst Plas. He joined the H.W.A. in July 1933 and thus had considerable experience. He had been living under an assumed name and informed the author that he was the first allied officer who had interviewed him. He was cooperative and appeared to be most efficient and it is thought that the information given by him is complete and trustworthy.’

Sunday, November 17, 2013

The unfortunate Henry W. Antheil and the State Departments strip cipher

During WWII the high level cryptosystem used by the US State Department was the M-138-A strip cipher. Unfortunately for the Allies this system was regularly solved by the codebreakers of several Axis nations.

I’ve given details of the German work on the strip cipher here and here.
However there is one thing about this affair that still bugs me. The German solution of the strip system was facilitated by the material they received from their Japanese allies.

Agents of the Japanese Military Police were able to enter the US consulate in Kobe in late 1937 and they copied the 0-1 ‘circular’ set of alphabet strips. This was used for communications between embassies and for messages from Washington to all embassies. This material was shared with the Germans in 1941.
However this material was not the only set of strips that the Germans were able to acquire covertly. Dr Wolfgang Franz, who was responsible for the strip solution at the German High Command’s deciphering department – OKW/Chi, said in his report TICOM DF-176, p6:

‘Especially laborious and difficult work was connected with an American system which, judging by all indications was of great importance. This was the strip cipher system of the American diplomatic service which was subsequently solved in part. After I had been working on it for a long time and was beginning to get some insight into the system, the work was greatly furthered by some captured material. This was given to me with no word as to its provenance.  From inscriptions and notes, however, one could infer that these were Japanese photographs. These were the basic material of the so-called ‘intercommunication strip cipher system 0-1’ and three further sets for special circuits between the Department and Reval, Tallinn and Helsinki (?) with designations of the type 19-1 or something similar. With these, several older messages could be read and the door was opened for further study of the system.’
According to David Kahn in ’Finland's Codebreaking in World War II’ in ‘In the Name of Intelligence: Essays in Honor of Walter Pforzheimer’:

The Finns got their break into the strip system when the German military espionage agency, the Abwehr, whose chief, Admiral Wilhelm Canaris, was a friend of Hallamaa, gave them photocopies of instructions for the strip cipher and of the strips for Washington's communications with the posts at Riga (which had been closed since June 1940) and Helsinki, as well as the 0-1 set.
How did the Germans get hold of the strips for Riga, Tallinn and Helsinki? From what I’ve read the Japanese were the source for the 0-1 set not the rest.

The strips from the embassies of the Baltic countries could have a connection with a Finnish civilian plane shot down by the Soviets in 1940.
The Kaleva was a Finnish civilian airliner that was shot down by Soviet planes on June 14, 1940, while en route from Tallinn to Helsinki.

Onboard was a US diplomatic courier, mr Henry W. Antheil, Jr who was apparently carrying diplomatic pouches from the U.S. legations in Tallinn and Riga. According to the wikipedia page the plane crashed at sea and the first on the scene were three Estonian fishing boats. Then a Soviet submarine reached the location and recovered all the material from the Estonians. This amounted to:
about 100 kg of diplomatic mail, and valuables and currencies including: 1) 2 golden medals, 2) 2000 Finnish marks, 3) 10.000 Romanian leus, 4)13.500 French francs, 5) 100 Yugoslav dinars, 6) 90 Italian liras, 7) 75 US dollars, 8) 521 Soviet rubles, 9) 10 Estonian kroons. All items were put on board of patrol boat "Sneg" and sent to Kronstadt

So the question remains. If the alphabet strips from the Baltics were recovered from the Kaleva plane, who got them and how did they end up in German hands?
Perhaps the Estonian fishermen were able to search the diplomatic bags and they retrieved the cipher material. Then when they got back to Estonia they could have given these to the military authorities who in turn shared them with the Germans.

That’s one theory.
Another one could be that the Soviets after recovering the diplomatic bags searched them thoroughly and recovered the alphabet strips. If that was the case then how could the Germans have gotten hold of them?

Could there be an exchange of secret material between the intelligence agencies of Nazi Germany and the Soviet Union? In the period 1939-1941 they were officially allies’….
Quite a mystery!

Saturday, November 16, 2013

Estonian signals intelligence service

The signal intelligence agencies of small nations usually receive little to no attention from historians, mostly due to the lack of primary sources.

The Estonian sigint agency monitored Soviet traffic during the 1930’s and cooperated with the similar departments of Germany and Finland. Unfortunately it is very difficult to find information on their operations and successes.
Some information is available from the very interesting article ‘Estonian Interwar Radio-Intelligence’ by Ivo Juurvee (Baltic Defence Review No. 10 Volume 2/2003) , uploaded on site

Some quotes:
‘The Estonian pre-war military intelligence service - the Second Department of the General Staff - and especially its radio-intelligence branch, Section D, have not been researched much…’

‘The Wireless Station of the General Staff in Tallinn intercepted the first radio messages of the Red Army during the War of Independence (1918-1920).’

‘In contrast to other parts of the Second Department, the personnel of Section D as of summer 1940 is precisely known: it was 26 people . two officers, 23 NCOs and one private. Nobody had been hired before 1936. This confirms the supposition that Section D was formed in 1936-1937. The second officer, Olev Õun, was taken to service only in March 1938; so far Andres Kalmus had managed to supervise the section alone. Radio-intelligence had gone through two major enlargements. The first of them was at the beginning of 1937, when Section D had just started its work. The second occurred in summer of 1939, when, according to President Konstantin Päts. secret decree from July 10, .due to complex situation [in Europe] naval radio intelligence has been reinforced.. With the order of the Commander-in-Chief General Johan Laidoner from July 22, the radio crew of the Second Department was enlarged .substantially..’
The top codebreakers were Andres Kalmus and Olev Õun. Note that these names also show up in some TICOM reports.

‘Captain Kalmus had followed military radio courses abroad.’
‘Olev Õun was especially talented, who was, in Hallamaa’s opinion, a phenomenal decipherer and had managed to break the latest code of the Red Army during the Polish campaign in September 1939.’

‘In 1939-1940 Section D units were stationed in Merivälja (7 km to the East from the city centre of Tallinn, probably next to the lighthouse of Viimsi, where the post of Naval Communications was situated, or somewhere in the area of nowadays Ranniku Road or Mõisa Road), Narva (probably at Olgino Mason 5 km to the North-East from city centre) and Tartu (probably in some of the units of the 2nd Division).’

‘When the Second Department closed down, it handed 51 items of literature over to the Red Army, including nine items concerning cryptology, a Russian-Estonian military dictionary and three Krypto ciphering clocks.’

Thursday, November 14, 2013

The British railways code

During WWII all the participants had some success in intercepting and decoding the radio traffic of enemy military units. Another type of traffic that proved to be very important for military operations was the traffic of the railways organization. By monitoring the movement of troops and supplies it was possible to identify the buildup of troops at specific areas of the front and thus anticipate enemy movements.

The codebreakers of Bletchley Park attacked the traffic of the German Railways - Deutsche Reichsbahn and started solving messages of the Eastern European network in 1941. Through this traffic they were able to monitor the movement of men and supplies to the East.
The German Army’s codebreakers were able to solve the code used by the NKVD railway troops and thus they also got information on the movement of supplies and the concentration of forces in specific areas of the front.

I’ve mentioned in my piece on German intelligence on operation Overlord that the Germans were able to solve the code used by railway troops in Britain in late 1943.
According to ‘Delusions of intelligence’, p46:

‘This same Heer station had broken into the British railroads codes by late November 1943 and claimed a 98 percent success rate in reading the two thousand plus signals produced by twenty-six keys in December 1943. Although not considered vital in peacetime, such intelligence on Britain proved important by providing information on the movement of troops and supplies.’
Obviously the solution of this traffic could have compromised the security of operation ‘Overlord’. More details on this system are available from the war diary of Inspectorate 7/VI and the reports of NAAS 5 (Nachrichten Aufklärung Auswertestelle - Signal Intelligence Evaluation Center). This was the cryptanalytic centre of KONA 5 - Signals Regiment 5, covering Western Europe.

The war diary of Inspectorate 7/VI shows that the radio traffic of the railways was first investigated in late August 1943 and in September a report was issued giving some information on these networks. There were two main networks, The one from South London, covered the territory of the Southern Railway (SR) and the Great Western Railway (GWR), the one in North London covered the area of ​​the London, Midland & Scottish Railway (LMSR) and the London & North Eastern Railway (LNER). Most of the traffic was from the first network and a few of the station callsigns were identified (Ashford, Tunbridge Wells, Chatham, London, Horsham). Some of the reports dealt with ‘coal positions’, ‘crippled wagons’, the removal of ‘rubble’ and cement shipments.
Investigations continued in October and in November they succeeded in solving the cipher used for station names. This was a paired Caesar, meaning the well known Playfair cipher. The square was changed each day and during the month 12 keys were solved. The results were communicated to NAAS 5 so that they could take over the solution of this traffic (called ECr27 in the reports).

In December ’43 a list of the frequent abbreviations and aliases appearing on the ECr27 was prepared and sent to NAAS 5. 

The reports of NAAS 5, E-Bericht 1/44 and E-Bericht 2/44 state that the code used by the radio network of the railways (Engl. Eisenbahnfunknetz) was solved almost completely in December 1943 and January 1944.

In December ’43 26 ‘keys’ and 2.304 messages were solved.

In January ’44 24 ‘keys’ and 1.871 messages were solved.

However in February ’44 the code was changed and from 16 February no such traffic was intercepted.

The solution of this traffic in the period that the Anglo-Americans were preparing the invasion of Western Europe may have given the Germans clues about the concentration of forces in the Southern areas of the UK. 

Sources: Delusions of intelligence, E-Bericht NAAS 5, Kriegstagebuch Inspectorate 7/VI

Monday, November 11, 2013

Solution of prewar Polish diplomatic code by OKW/Chi

In the field of signals intelligence and codebreaking Poland, despite being a small state, distinguished itself by being the first country to solve messages enciphered with the German military’s Enigma machine.

However the Poles did not have similar successes in the field of crypto-security. Their diplomatic, intelligence and resistance movement codes were regularly read by the Germans prior and during WWII.

An interesting case is the solution of the main Polish diplomatic code by the codebreakers of the German High Command’s deciphering department – OKW/Chi, during the 1930’s.

Details on the Polish code are available from TICOM report DF-187G, pages 11-19. This report was written by Wilhelm Fenner, head of the cryptanalysis department of OKW/Chi.
According to Fenner the Polish code used since the 1920’s was 4-figure. Through repetitions in the code values the Germans deduced that this code was enciphered with a simple substitution of the digits. Obviously this system offered limited security. Simply by comparing each day’s most frequent code groups it was easy to figure out the daily substitution.

Another serious mistake made by the Poles was that the substitution table for the month was not created randomly but instead had systematic features that helped the Germans in recreating them.

Later on the substitution system was replaced with a more secure additive system. Again however the Poles made the mistake of taking half measures. The additive sequences used to encipher the 4-figure code were too short, and they were used for long period of time. This led to messages being enciphered with the same sequences and these ‘depths’ could be exploited by OKW/Chi.

Only during the late 1930’s was the security of this system upgraded by using long additive sequences and having different enciphering tables for incoming and outgoing traffic. Of course one can argue that by then it was too late to make a difference.

During the war the Poles continued to use additive sequences but these were read by the Germans. This however doesn’t mean that these systems could be exploited at will by them. Instead it was necessary to intercept as much material as possible and to use special cryptanalytic equipment.


Thursday, November 7, 2013

Operational research in Northwest Europe - No. 2 Operational Research Section

A very interesting report is available from site dtic online. This is the report Operational research in Northwest Europe , the work of No. 2 Operational Research Section 21 Army Group.(originally found through world of tanks forum user GhostUSN)

The No2 research section teams followed the Allied ground troops and estimated the performance and effectiveness of Allied weapons and tactics by gathering data from the battlefield.

There are separate chapters for airpower, artillery, tanks and infantry weapons.

Saturday, November 2, 2013

WWII Myths – German tank strength in the Battle of France 1940

In May-June 1940 Germany shocked the world by defeating the combined forces of France, Britain, Holland and Belgium in the Battle of France.

At the time no one expected that the French forces would be defeated in such a short campaign. During the interwar period the French Army was thought to be the best trained and equipped force in Europe. On the other hand Germany had only started to rearm in the 1930’s.
The sudden collapse of France led to a search for the reasons of this strange defeat. There was no shortage of excuses. Every part of France’s defense strategy came under attack, from the old Generals of WWI that tried to control the battle from the rear to the funds wasted building the Maginot line.

General Gamelin who commanded the French forces told Churchill that the defeat was due to: ‘Inferiority of numbers, inferiority of equipment, inferiority of method’.
Was that true? Considering the role played by the German Panzer divisions in cutting off the northern part of the front it is important to have a look at their strength.

Did the Germans have more tanks than the Franco-British Alliance?
According to Panzertruppen vol1, p120-121 the German Panzer divisions used in the Battle of France had the following strength on May 10 1940:

Pz I
Pz 35
Pz 38
Pz Bef
1 Pz Div
2 Pz Div
3 Pz Div
4 Pz Div
5 Pz Div
6 Pz Div
7 Pz Div
8 Pz Div
9 Pz Div
10 Pz Div

The same source gives the following losses at the end of the battle in page 141:

Pz I
Pz 35
Pz 38
Pz Bef

How did the German tank strength compare with the Allies? According to The Blitzkrieg Legend: The 1940 Campaign in the West, p37-38 the French Army had in the Northeastern Front 3.254 tanks, the British Expeditionary Corps had 310 plus 330 in transit from the UK, the Dutch Army had 40 armored vehicles and the Belgian Army roughly 270. Total for the Allies came to 4.204.

So in the field of tanks the Germans were definitely outnumbered. If we look at tank types it’s easy to see that they were also outgunned. Their main vehicles were the Panzer I and Panzer II. The first had only two machineguns and the second a 20mm gun. Against Allied tanks equipped with guns of 37mm caliber and over they were cannon fodder.

The German victory was not due to a numerical or qualitative superiority in armored vehicles. Instead it had to do with the way they used their armored forces, grouping them together, supporting them with ample airpower and providing them with dedicated infantry, anti-tank, artillery and communication units.