|The Biafran Ciphers|
The idea to this release comes from different inputs. One is GCHQ’s 100 years anniversary and their promised release of many post-war documents in conjunction with the publication of their authorised history written by Professor John Ferris entitles “Behind the Enigma: The Authorized History of Gchq, Britain's Secret Cyber Intelligence Agency“. This made me realise that my Crypto Cellar now is more than 50 years old and that perhaps I also was sitting on documents of some historical or curiosity value. That the selection became the Biafran radio messages containing a large number of cipher messages is due to my good friend Jan-Olof Grahn’s publication of the third volume of his books on Swedish signal intelligence, “Om svensk signalspaning: Kalla kriget” — “On Swedish signal intelligence: The Cold War.” Here Jan-Olof Grahn has a section on the Nigerian Civil War, also known as the Biafran War, running from 6 July 1967 to 15 January 1970 or 925 days. He reveals that Biafra had a radio telex or radioteletype (RTTY) link from a station in Portugal, called Biscaia, to a station in Biafra, called Luanda/3 or simply LDA/3. However, perhaps the most surprising is that the Swedish signal intelligence agency, Försvarets radioanstalt — FRA, was following this traffic from late summer 1967 until the last day the link operated on 30 January 1970 and that they were breaking the cipher messages and reading the traffic. They first detected the station Luanda/3 operating on the frequency 17320 kHz with teleprinter messages in the international teleprinter code ITA2, also known as the Baudot/Murray code, with a speed of 50 Baud. A few days later the other end of the link, the station called Biscaia or BIS, was discovered operating on the frequency 20877 kHz. Direction finding showed that the stations were indeed operating in the direction of Luanda in Angola and in the region of Biscaia or northern part of Spain or Portugal. Later information from the telegram traffic made FRA believe that LDA/3 was situated inside Biafra and BIS most likely was situated in or close to Lisbon in Portugal.
The messages published here are not from FRA, they no longer have copies of the raw traffic and the messages published by Jan-Olof Grahn in his book comes from me. The background story of how I obtained these messages in August and October 1969, exactly 50 years ago, I will leave for later perhaps as a posting on my blog Crypto Cellar Tales. What I can say is that the messages were received in Norway in a similar way to how FRA received the messages, however both the receiving equipment and the antennas were probably much inferior to the FRA equipment. The LDA/3 was spotted a few times but the signal was too weak for reliable reception and therefore all the messages are from the station in Biscaia, BIS.
Unfortunately also my collection of this material has been unintentionally weeded due to many removals and also a case of an accidental flooding of a cellar storage place. What is published here is all what is left. What I know is missing is the radio log with all the notes and also several pages with operator chats from the Biscaia station. The few notes that are left shows that the messages were received on mainly three frequencies: 19270 kHz, 20890 kHz and 23800 kHz. The last frequency has the date 19 October 1969 which might indicate that it was only spotted on this single day. What I remember is that sometimes the stations would operate close to but not exactly on these two main frequencies. The reason for this was probably due to disturbences from other services in that range of frequencies. Most likely the stations were operating illegally in the sense that they used frequencies not allocated by international or national agreements. The Luanda station using the call LDA/3 looked initially as a callsign allocated to Norway and I therefore reported this to the Norwegian Telecommunication Administration with copies of some of the traffic. However, they did not seem duly concerned and they only asked me to inform them if I was able to establish the exact locations of the Luanda and Biscaia stations, something I never succeeded in doing.
Publishing cipher messages of unknown content always brings up the issue of the sensitivity of the hidden plaintexts. If the messages are broken will their content create any public offence? In September and October 1978 I told Brian J. Winkel, who was then the chief editor of Cryptologia, about the messages and I gave him copies of a few of them. The idea was to publish an article about the Biafran ciphers but not necessarily release any of the messages. However, first it was necessary to break the ciphers and there I had failed. In the end we decided not to publish anything because it seemed at that time impossible to write anything sensible without first breaking the ciphers or publish the ciphers for others to break. Now 40 years later and 50 years after the messages were intercepted I think it is safe to release it all. The history of the Nigerian Civil War is now well known and so are most of the actors. The US Department of State has released most or possibly all of their documents from the conflict, including memoranda, situation and intelligence reports, and State Department telegrams in their collection Nigerian Civil War. The only exception are some pages marked Top Secret — Umbra, which show they contain signal intelligence information.
I do not think the messages are of any real historical importance. They are more of a curiosity, however the radioteletype link which connected Biafra with the outside world and their many roving representatives, ambassadors and envoys show the sophistication of the rebellion state Biafra. The radio link has only briefly been mentioned before such as in John J. Stremlau’s book “The International Politics of the Nigerian Civil War”, where it is called the Lisbon telex link, and of course now in Jan-Olof Grahn’s book “Om svensk signalspaning: Kalla kriget.” Biafra also made their own currency as can be seen from the one pound note shown below and they also made their own postal stamps. The Igbo population was and is highly educated and many of their people were professors and graduates with PhDs. This is also reflected in the sophistication of the Biafran state, its approach to the civil war and its highly developed international diplomacy. The radio link played a part in this and it is suspected that the messages will contain information about the relief operations, but also the purchase of arms and other military supplies as well as reports to and from their Foreign Office people. However, Jan-Olof Grahn writes that some of the messages had more personal content such as information about purchases of luxury goods like alcoholic beverages and clothes.
The purpose with the release is to throw some more light on this rather obscure radioteletype connection from Biafra to Lisbon and to present the many eager amateur codebreaker out there with yet another challenge. It is my hope that someone will succeed in breaking at least the transposition cipher that is used in the majority of these messages. I also hope that somebody who is acquainted with the radio link and the stations in Biafra and Biscaia is willing to come forward and tell us something about the equipment that was used, exactly where the stations were situated and perhaps some of the operational details. We know that FRA deciphered and followed this traffic and it is therefor logical to believe that NSA, GCHQ and the signal intelligence agencies of France and Germany also broke and read these messages. Other interested parties might have been Portugal and the Soviet Union that also took part in this conflict. Future releases of documents from these agencies might one day give us the full picture.
Biafran One Pound note.
Photo: With permission of Jan Olof Grahn, Sweden © 2019
Biscaia testing to LDA/3.
Here are cipher- and cleartext messages from the Biafran radio link.
I have received questions about transcripts of the Biafran ciphertexts. No such transcripts
have been planned, however if people are willing to send me transcripts then I will try to
publish them here.
A few message have now been transcribed. Teleprinters do not have upper case, the letters are either lower or upper case depending on the teleprinter model. Teleprinters only have letter and figure shift changing between letters and figures and punctuation. For the transcripts I have decided to always use upper case letters even if the facsimile copies mostly show the messages in lower case.
The 5-figure messages will be added later after the 4-letter traffic has been transcribed.
The radio messages contain both clear- and ciphertexts. The ciphertexts in this collection are of two types, messages in groups of 5-letters and 5-figures. In 1969, shortly after receiving these messages, I made a quick analysis of both types of messages. A frequency analysis of the 5-letter traffic showed a monoalphabetic distribution corresponding to English plaintext, and the form of the messages made me believe that they were enciphered by a transposition system. The messages are divided into a number of blocks that often, but not always, consist of an odd number of groups. The message BAL025 from 04 August 1969 is divided into four blocks marked 71/4, 142/4, 11/4 and 22/4. The number behind the slash is the date when the message was enciphered. It therefore seems logical to believe that the same or a similar transposition is performed on each of these blocks. In 1969 and also later in 1978 a few feeble attempts were made to break into this system. Irregular columnar transposition and some other of the more common transposition systems were investigated but without any success. In the end I decided that either double transposition was used or a kind of stencil system. The distribution of the blocks of all of the messages at the various dates are given below.
Transposition Block Distribution 04 August 1969 11/4 - 22/4, 32/4 - 63/4, 33/4 - 65/4, 61/4 - 122/4, 71/4 - 142/4 06 August 1969 31/6 - 62/6, 71/6 - 142/6 09 August 1969 56/9 - 112/9 10 August 1969 64/10 - 128/10, 67/10 - 133/10 11 August 1969 75/11 - 149/11 15 August 1969 54/15 - 107/15, 75/15 - 149/15 (DARTLX 7 of 15/08 sent as BAL141 on 16/08) 64/15 - 128/15, 75/15 - 149/15 (DARTLX 9 of 15/08 sent as BAL139 on 16/08) 16 August 1969 43/16 - 86/16, 75/16 - 149/15 19 October 1969 36/19 - 72/19, 41/19 - 82/19, 71/19 - 142/19 20 October 1969 21/20 - 42/20, 28/20 - 55/20, 43/20 - 85/20, 71/20 - 142/20 21 October 1969 20/21 - 40/21, 40/21 - 80/21, 35/21 - 69/21, 71/21 - 142/21
From this it seems most likely that for each day there is either one or a few stencils of various fixed sizes that are being used. If there is one single stencil being used then the number of letters in each block will be marked and delimited on the stencil before the plaintext is filled in. If stencils of various fixed sizes are used the appropriate size will be selected and then again the appropriate area delimited to fit the number of plaintext letters.
In his book Jan-Olof Grahn has an overview of the content of the messages that FRA deciphered for the period 1967 to 1970. The 9th October 1969 Colonel Ojukwu reports that a number of stations have received cipher pads for intermission communications, which they shall start to use from 15th October. This can sound like one-time pads but it can also mean special stencils for the transposition system.
The 5-figure system is most likely a code but it could also be a substitution cipher. No detailed analysis has been made of this system largely because there are so few of these messages. Jan-Olof Grahn reports that the ciphertexts that FRA deciphered consisted of a transposition cipher which was later complemented with codes and the use of an electromechanical cipher machine. None of the ciphers in this collection seem to be machine ciphers.
How many telegrams FRA intercepted and processed is not known, but in 1967 they reported 143 telegrams to the Swedish authorities, 397 telegrams in 1968, and 688 and 12 telegrams in 1969 and 1970 respectively. The first messages were most likely broken by Bengt Beckman (Unfortunately the English Wikipedia version is totally wrong), however the work was taken over and continued by Bojan Frykholm, FRA’s specialist on transposition ciphers.
From about 1 September 1967 the Biafrans also started to use codewords in the ciphertexts for place and country names. This list of codewords was replaced with a new and expanded list in September 1968 and which was in use until the end of the war. In this new list frequently used names had multiple codewords, e.g. Biafra had the codeword Annseh in the first list while in the second list the codewords were Dalfon and Flandin.
On 15 July 1969 Chiji reports from Paris that it now exists clear proofs for the fact that the “old code” no longer is secure. According to Dr. Dike (Dr. Kenneth Dike) the “old code” apparently was broken 18 months ago. If this refers to a code or the transposition cipher is not clear from what Jan-Olof Grahn writes, however it seems that from then on a new system was taken into service. Colonel Ojukwu replies to this on 16 July, with address to all Biafra units, that special instructions are given to the cipher operators units that from now on messages concerning supplies, security and political reports must be sent with the new system. The only exception is for circular messages.
Links to other sources about Biafra and the Nigerian Civil War:
I am interested to hear about errors, typos, outdated information or any other information you might have on the Biafran cipher or radio messages. And of course I am interested in hear from anybody who has managed to break any of the ciphertexts.
You can contact me here:
IN THE NEWS:
All rights to the material described on these pages belong to
Frode Weierud, © November 2019
|Back to Frode’s CryptoCellar||