|
Home Page The Red Home Page
|
Beginning of the Use of Cipher Machines by Japan (1927-1932)Preliminary note: Much of this section is inspired by Tomokiyo's webpage that talks about RED cipher machine (see References). Between the two world wars, many countries, including Japan, considered acquiring cipher machines to replace codes, which were very complex to use. They required trained personnel, and the codes were time-consuming to perform. It was in 1927 that Kakimoto Genichiro of the Japanese Navy conceived the idea of making a cipher machine. Several prototypes were tested at the London Naval Conference in January to April 1930. Sakuma, Chief of Telegraph Section of the Foreign Ministry, in 1931, asked for Cabinet approval to introduce the Navy's new cipher machine for use during the Geneva Naval Conference scheduled for 1932. The encryption time, divided by four compared to the use of codes, argued in favor of the use of cipher machines. Note: The first Japanese cipher machines were very insecure. But to Japanese cryptologists at the time, they seemed very secure, even more secure than codes. The key length, over 2000 characters, and the astronomical number of plugboard configurations were misleading. In 1932, two machines were officially adopted, based on the same principles:
Operation from 1933 to 1939Start of OperationThe start of use of the RED machine is unknown, but the first American interceptions date back to March 1933. Abandonment of Vowel/Consonant SeparationIn March 1934, the Japanese abandoned the separation between vowels and consonants but maintained a separation between six letters (Sixes) and the other twenty letters (Twenties). One letter from each group was necessarily encoded by another letter from the same group. 1937: The Plugboard configuration changed each dayIn 1937, the Plugboard configuration changed daily. Previously, it was valid for a period of 10 days. This event did not bother the SIS cryptanalysts. Indeed, the Plugboard configurations followed an algorithm (see References, the book by Deavours & Kruh). 1938: Machine improvementOn 1 December 1938 the RED machine was modified by the addition of three special commutators of interrupted motion to encipher highly secret messages. The US Navy calls this model M-3A. 1939-1941: Replacement by the Purple machineThe RED machines are old and in poor conditionAccording to the SIS, in 1939, only one of the two machines owned by each embassy was in working order. In some embassies, both machines were in poor condition. This poor condition was manifested by encryption errors. Many messages had to be retransmitted. Announcement of the replacement machineThe cryptanalysts monitoring the RED traffic learned, while decrypting the messages, that a new machine (Angooki Taipu B) would replace the RED machine (Angooki Taipu A). They also learned that a specialist would visit the embassies to get the new machines working. The first message encrypted by the new machine (called Purple by the Americans) was sent on March 20, 1939, between the embassy in Warsaw and Tokyo. Gradually, the Red traffic diminishes and is replaced by Purple traffic. All important messages are then encrypted with the new machine. Red machine traffic stopThe last RED message transmitted is dated August 21, 1941. CryptanalysisThe RED machine was broken by several cryptographic agencies: The US Army (SIS)Rowlett & Kullback, under the direction of Friedman, broke the RED machine in 1936. The following page describes their exploit in detail (link). The US NavyIt appears that it was Agnes Driscoll who, around the same time as the SIS, rebuilt the RED machine. It seems that she was largely aided by knowledge of the Orange machine, the operation of which was discovered through espionage. Ladislas Farago (see References) describes in detail this operation, which took place in July 1935. The naval attaché's office in Washington were inspected by fake electricians who intervened following a power outage (caused by the Americans). Obviously, the Orange machine was inspected in detail, with Jack S. Holtwick being tasked with reconstructing the machine. The BritishIn the 1930s, Hugh Foss and Oliver Strachey cracked the Orange machine used on the Japanese naval attaché network. The word Oyobi (and) was very useful to cryptanalysts because it consisted of four vowels, three of which were consecutive. Shortly after, in 1934, the British also cracked the RED machine, which used the letters Roma-ji instead of the Kana syllables. A machine emulating the Japanese machine was built by Kenworthy of the Metropolitan Police Wireless. It was able to decipher the many RED messages. The GermansIn the spring of 1936, Kunze, of the German "Pers ZS" (the agency responsible for attacking diplomatic cipher systems), attacked the Orange machine and subsequently cracked it. In 1938 (and until 1939), the Pers Z S section, working with OKW/Chi, broke the keys of the RED machine daily. SIS's Study of Japanese Diplomatic CiphersBefore tackling the RED machine, Friedman's team, starting in 1933, studied Japanese diplomatic codes. First, they relied on the archives of the "Black Chamber" that SIS had inherited after the closure of Yardley's office. They were then able to study current traffic. In March 1934, the J-6 code was cracked. It was not very important in terms of message content, but knowing it allowed Friedman's team to familiarize themselves with the message format, vocabulary, headers, etc. All this knowledge was very useful to Rowlett and Kullback when they tackled traffic encrypted with the RED machine. Between 1933 and 1935, five Japanese diplomatic codes were used in rotation at three-month intervals. These systems were cracked. They consisted of two and four letters. By 1938, nine codes were regularly read, although they were more complex than before the publication of Yardley's book (The American Black Chamber) in 1931. These codes were widely used by smaller embassies. Indeed, cipher machines (RED and Purple) were only installed in large embassies (London, Paris, Washington, etc.). If non-vital information needed to be sent to all embassies, it was done so by codes. Note: Rowlett's memoirs not only contain the exploit of breaking the RED machine, but also discuss his work on Japanese diplomatic ciphers. ReferencesBooks and Articles
Internet
|