German Army cryptographic systems of World War II

The history of the preparation, printing and distribution of cryptographic keys by the German Army was described by Walter Fricke and Erich Hüttenhain.[14] According to Fricke, the keys for the Army were produced at the beginning of the war on In 7/IV's own press. The number of personnel required was small, about 20 people in all, since only enigma, weather and book staple keys (German: HeftSchlüssel) were printed.

The number of personnel required for producing keys increased in 1940, when the Heftschlüssel was replaced by the double transposition and in 1942 when the double transposition was replaced by the TS-42a and the NS-42. A considerable increase in the personnel took place in the section as a result of the introduction of the key tables to the W/T tables. With the introduction of the SZ 40 and SZ 42 and other cryptographic systems which required keys, the amount of work entailed in the preparation of keys became too great for In 7/IV alone. In 1942, therefore, the production of key table manuscripts was transferred to the IBM section of the unit, and the printing to the Reich press in Berlin. In 7/VI was confined to reading proof and distributing the finished product.

At the time of the bombing attacks on Berlin on 22 and 23 November 1943, the printing press and all key material of In 7/IV at the Tirpitzufer 76 was destroyed. The key producing section was then moved from Berlin into emergency quarters prepared some weeks before in the Army Signal Academy located in Halle (German: Heeres Nachrichten Schule) During this period, private printing firms were increasingly drawn on for the production of keys, first, because the Reich printing press could not meet the requirements, and secondly, decentralisation was becoming increasingly necessary to avoid bombing. About 20 firms in the central German area were given contracts. The high number was attributed to the introduction in 1944 of the Stencil System 44, the Rasterschlüssel 44.

In May 1944, the key production section was transferred to the Hindenburg barracks at Dresden owing to a lack of space. This situation remained until 1 November 1944, when all the tasks of producing, printing and distributing of keys were transferred to Section IIIa of the OKW/Chi. This section was established in Dresden in the HQ in which the key preparation section of In 7/IV had occupied. In March 1945, Section IIIa was transferred from Dresden to the Army Signal Academy located in Halle due to the approach of the Soviet Army. Work was never begun, and on 12 April, keys and material production machines were loaded onto three trucks to be sent to southern Germany by goods train. Keys left in Halle were destroyed by Germany at the approach of the United States Army.[15]

Before 1939, the Army High Command had no security organisation of its own. Questions of security concerning Army systems were transferred to the Codes and Ciphers section of the German Defense Ministry. Erich Hüttenhain stated that he and Fritz Menzer conducted a security study on the initial model of the SZ40 and found that it could be solved in two days.[8] This led to its improved form.

In 1939, the Army High Command established its own Signal Security Agency (In 7/IV) which functioned as a unit until 1942. During the period of its operation, In 7/IV' examined the plugboard Enigma[16] and made security studies on two other cryptographic machines proposed for Army usage, the M-40, first proposed by Fritz Menzer and the SG-41.[17] The initiation of the security study of the plugboard Enigma was occasioned by Der Fall Wicher or case Wicher, the suspicion that Polish cryptanalysts of the Biuro Szyfrów had already read Enigma machine traffic by 1939.[18] It was known that several cryptanalysts including Friedrich Böhm made security studies of the machine, and found it to be secure, although it was insecure. The Army continued to use the system, until Walter Fricke[19] informed the Army that the Armies manner of using the Indicators led to an easy solution. At his recommendation, the indicator system was changed.[20][21]

The two machines proposed for Army use, the M-40 and SG-41 were invented by Fritz Menzer, who worked for OKW/Chi. Security studies on the M-40 were conducted by Wachtmeister Heinrich Döring and Otto Buggisch.[22] Buggisch stated that the studies proved the device to be moderately secure, but that it was never used because it was as bulky as the plugboard Enigma but could not print letters.[23] With regard to the SG-41, the studies made by OKH/Chi showed it to be superior to the M-40, but regarding its use, Buggisch stated:

The Army hemmed and hawed and never did adopt it[22]

In 1942, the security testing unit and most of the testing personnel of In 7/IV were transferred from their former unit to the Group IV of OKH/Chi Hans Pietsch, S Steinberg, Herbert von Denffer, Hilburg and Hans-Peter Luzius were named as those transferred.[24] Security work was the responsibility of Group IV, Referat I, Section Ib, Subsections 7 and 13. Subsection 7 worked on testing German hand systems, 13 on machine systems.[24] The first studies of Subsection 7 on German hand systems proved that all systems currently in use by the Army were solvable. As a result, OKH/Chi was ordered to collaborate with OKW/Chi in the development of new systems for the Army.[25]Although OKH/Chi would have preferred to establish this section within itself, where the preparation of systems would be done in close cooperation with cryptanalytic specialists, this point of view was not recognised by the Army and OKH/Chi was ordered to send mathematicians back to In 7/IV. The mathematicians sent were Fricke, Jesse and Kehren.[25] From that time on, subsection 7 only developed hand ciphers handed to it by the Field Army. Mettig noted that the amateur systems with which the section dealt were very bad and betokened great ignorance on the part of the Field Army in regard to code and cipher security.[24]

Whenever the German Army was asked to change a system, there was always a storm of protest. It was not them but the staff of General Fellgiebel (Chef HNV) (German: Heeresnachrichtenverbindung) (HNV) which made the decision on which methods to be used. The results depended on whether the officers at HNV at the time, happened to know anything about cryptography. He usually did not, as it was a specialised field. In 1942, all hand systems were solvable. When the HNV was told this, the reply was that Germany had won all her battles so far, using these systems, and there was no need to overload the troops with new methods.[9] Only with the greatest difficulty was the Field Army persuaded to change its methods.

One of the specific ways in which the Field Army consciously hampered progress in security studies was to refuse to furnish OKH/Chi actual traffic. Fricke stated that OKH/Chi never knew how the Field Army actually used the systems which it approved. When OKH/Chi asked for studies, it was given specially prepared messages such as We are standing in Berlin and See the Polish infantry coming down the Frankfurt Allee.[9] However, the Field Army made a brief attempt in 1941 to provide the unit analysts with actual traffic. For this purpose a Signal Intelligence Regiment of the Replacement Army (German: Nachrichten Aufklärung Abteilung/Chef der Heeresrüstung und Befehlshaber des Ersatzheeres) was formed. While two companies of this unit were to act as administrative units for personnel of In 7/VI, the department of the Army Signal Security agency responsible for cryptanalysis of non-Russian traffic, the third company was an intercept unit in the field collecting material for the analysts. Owing to personnel shortage, this unit was dissolved in February 1942, and no subsequent attempt was made by the German Field Army to procure actual traffic for OKH/Chi. [9]

As described above, it was difficult to persuade the Field Army to accept as valid the security studies made by the analysts of In 7/VI. Not until a cellar that contained equipment for tapping land lines used by the T-52c teleprinter, was raided in Paris in early 1943, was the Army High Command persuaded that the security studies made by the OKH/Chi had been valid.[27]

The attitude of the Field Army could be traced largely to the ignorance on all levels of matters pertaining to codes and cipher security. Traffic handed to In 7/VI for scrutiny betrayed the lamentable ignorance of the Field Army.[33] Signal tables set up by the troops revealed serious cryptographic errors such as the failure to change keywords for long periods.[34] In an attempt to correct this ignorance, OKH/Chi gave lectures at the Army Signal Academy located in Halle (German: Heeres Nachrichten Schule) and issued instructions on code and cipher security. The situation, however, was never satisfactory resolved. Buggisch stated, while researching the C-36 device that the studies made on the device by B. et al. were used by Oberinspector Kühn and Major Mang to forestall the introduction of the device into the German Army, as advocated by Major Mang. Buggisch stated later,

it was tragic-comic:When OKH/Chi detected an insecurity, it was not able to achieve effective remedies; if In 7/VI wanted to install new devices, it had even more difficulties. The Army "hemmed and hawed" and never got around to acting.[34]

German signal intelligence did not take a creative, non-linear view of attacking Enigma security, or any other secure system. When reporting on the security of Enigma or even of other ciphers including enemy systems, then focused on the statistics of cryptology. Their security investigations and cryptanalytic reports examined either the problems of physical compromise or of possible statistical methods for breaking systems. Among the surviving wartime documents, no reports deal with the more subtle ways around and into cryptologic problems.[35]

The Army had cryptanalysts attack their own Enigmas, and in these attacks Germany's experts seem to have stuck to statistical and straightforward approaches. They conducted statistical attacks on Enigma-enciphered traffic both with and without the help of captured documentation. Both the OKH/Chi and Pers Z S found methods to solve the original unsteckered versions of Enigma.[35] However, it was known they never tackled live traffic, for reasons described above, and instead stuck to straightforward attacks conducted under artificial conditions. All these examinations concluded that the Enigma would defeat enemy cryptanalysts. Throughout these investigations, German cryptologists seem to have never realised the true reduction of Enigma's probability space under real conditions, particularly those reductions arising from operational procedures and human error.[35]

German cryptanalysts had no luck solving Allied high-grade ciphers and this lack of success only reinforced their believe in Enigma strength and vice versa. As Enigma had proved unbreakable, the Typex would proved even more impervious to attack, and hence the Army did not expend much effort on attacking the device.[35]

Relying too heavily on theoretical statistics made Germany vulnerable. They knew that Great Britain had captured Enigma machines and associated documentation by the middle of the war. They also knew that Britain had a history of successful cryptanalysis and appreciated the abilities of British intelligence, yet they relied on a statistical security that took neither operational reality into account nor their opponent's capability. They ignored simple factors which could have hindered, if not defeated the Allies. One example, was the complete replacement of the rotors, by the Army, which was never done for various reasons.[35]

Germany did recognise many of these procedural problems and they frequently reprimanded Enigma operators for insecure habits. In additions, several agencies, conducted investigations into Enigma that hinted at the system's general compromise. However, only the examination of live traffic under real conditions could have suggested the true level of security for Enigma, or indeed any other system, and this was never done.[35]