eclassified and A xoved for Release by NSA on 10-17-2012 pursuant to E O 13526 MDR Case # f515 Bl JWVU GBPiJlB15W rI DWV lB15I DWlB15 lB WlUiJlDl f W l1U11I1ID WUJ'i VGBP JWUJI1 Brn I1UI l1l rn DECEMBER 1986 L TECHNOLOGY SEEPAGE U WHO IS SHE U o o o o o CRYSCOM UJ o o o o QUOTE WITHOUT COMMENT U THE LANGUAGE IDENTIFICATION PROBLEM U PRODUCT vs PROCESS UJ o CHANGING JOBS U o o ODE TO THE CA INTERN PROGRAM U A COMPOSITE CONVERSATION UJ BOOK REVIEW UJ o CONFERENCE REPORT UJ LETTER UJ OPPORTUNITY U BULLETIN BOARD UJ CISI CONFERENCE U THE STORY OF A PRIVATE GERMAN CIPHER U NSA-CROSTIC No 63 PLUS U o o IIAtdDlE li'IA SMitH I IANtdELS ONLY o 1 9 9 10 ----4 1 11 18 I 21 23 L -- --- J1 2 4 o 26 L I ---II 29 31 32 33 33 1 34 36 o Vera Filby I f I CLASSIFIED BY r SA EURSSM DEClASSll Y ON NOT RELEASABLE TO CONTRACTORS 86-36 123 2 OFi inatin A eAeY'5 DeterA liAatieA ReEtl lirea DOCID 4033692 Published by PI Techniques and Standards PUBLISHER o oo 1 -- l @ o 86 - 3 6 BOARD OF EDITORS EditOr I'-- Oll ctio ' f' J J L-- k963-3490 963-6211 To submit articles or fetters by mail send to Editor CRYPTOLOG Pl HQ 8A187 If you used a word processor please'include the mag card floppy or diskette along with your hard copy with a notation as to what equipment operating system and software you used via PLATFORM mail send to r y p t I 9 at barlc05 bar-one-c-zero-five note no 'o's in cry p t I g NOTE CHANGE Always include your full name organization and secure phone number I What this agency needs is a Julia Child to demystify personal computers 1 WHERE ARE YOU u l963-11 03 963-5877 Cor P JterSy tems 1 963-1103 Cryptanatysis 1 963-5238 Cryptolingtltstics 1 963-1596 ' J 1963-5292 Index Information$ciefl e 1 1 963-1145 InformationSe t Jrity Georae F Jelen 1859-1211b Intelligence I'lesear ch 1 1963-3845 Language 1 963-3057 f 963-5566 Mathematics J 96 -6430 Puzzles Science and Technology 1 968-8075 Vera R Filby 968-8014 Special Research Robert J Hanyok 963-5734 Traffic Analysis Illustrators JULIA CHILD December 1986 VOL XIII No 12 ' For Change of Address mail name and old and new organizations to Editor CRYPTOLOG PI HQ 8A 187 Please do not phone Consider what she did for French haute cuisine she brought it within the ken of any American who could read How Very simply she wrote a step-by-step handbook with ample illustrations and explanatory notes all in an informal joyous and encouraging tone Now there's a whole generation who never knew that French haute cuisine was once the exclusive preserve of a select group of the anointed What is odd about personal computers is that they were intended as the name rather suggests to be used by anybody It's not working out that way because the user's manuals are poor Scan the business pages of the local newspapers some time-you'll see that enterprising people have recognized this and are profiting from it by offering live-in seminars and other courses And at a price mghway robbers It would be advantageous to develop a handy-dandy for personal computers and maybe software as well to be issued by the PCIC along with the hardware and software As it is many terminals are unused or underused solely as word processors for lack of understandable instructions So in the interest of productivity-think of those hours spent grumbling as well as trying to cope-it behooves Somebody Up There to commission a cookbook-type handbook to personal computers And when it's published please send us a copy Contents of CRYPTOLOG should not be reproduced or disseminated outside the National Security Agency without the permission of the Publisher Inquiries regarding reproduction and dissemination should be directed to the Editor 1 01 OI I IGIAI Y J O LJPY DOCID 4033692 P L 86-36 _ _ _ _ _Ipll P L 86-36 December 1986 CRYPTOLOG page 1 SB6IW'P OT RELL' S BLE TO EOPHRAETORS EO 1 4 c IIA WU3 YlkOOMHtf' OII MHfBLS ONL'l L 86-36 DOCID 4033692 SHORE'f' P L 86-36 EO 1 4 c December 1986 CRYPTOLOG page 2 SB6R8'P r OT RELEASABLE TO COfHRAeTORS HANDLE VIA eOMHff' eIIA H ELS ONLY DOCID 4033692 FREQUENCY HOPPING VARIABLE DWELL' TIr tlES -'--A - - - - - - - - tl f jI r F R E Q u E N C - -- -- C3 -------- P L 86-36 EO 1 4 c - Y C3 1 - - r- 1111111111111111111111111111111I11111111111111111111111IIIllIIII TIME spreading code used by the' target radio is known substantially less power is needed to jam its signal If the spreading code consists of a relatively short recursive sequence recovery can be made by exhaustive hypothesis testing techniques Research to develop exhaustive methods of recovering short spreading codes would be unlikely to uncover sensitive algorithms Longer sequences such as those above the R20 level cannot be recovered quickly today by exhaustive testing To succeed against them algorithms that parallel those used in extremely sensitive cryptanalytic attacks would have to be developed he services' requirements are genuine and critical to their mission They do need a capability to intercept or to jam spread spectrum signals in a tactical environment There are some who question whether predictive jammers or non-exhaustive spreading code recovery systems are the best - or even viable - ways to provide these capabilities The services however assert that such equipments are vital to fulfilling their mission And they have the independence the will and the money to pursue development December 1986 POLICY ON CONTROLLING RESEARCH D There is no policy that requires the services or other government agencies to coordinate with us before initiating spread spectrum research Our knowledge of some projects was acquired by chance We have no way of gauging what portion of the total nonAgency spread spectrum research and development of a cryptologic nature these projects represent It would take a gigantic manpower-intensive effort to uncover and track all such projects and in the current climate it would be awkward to do it D Fortunately none of the research that has surfaced to date really qualifies as a smoking gun But as the number and sophistication of CRYPTOLOa page 3 SEeRE PWT RELEASABLE TO EOPHRAETORS IIAP'WL6 VIA GOMHff' CHAUUELS 0r4'Pt P L DOCID EO 4033692 Direct sequence spread spectrum signals are those for which the carrier is modulated by a digital code sequence whose bit rate is much higher than that of the information signal The application of o this spreading code spreads the signal across a broad bandwidth The intended receiver of the signal knows the spreading code and uses that knowledge to collapse the signal o back to its orginal form To intercept or jam the signal effectively the spreading code must be recovered The binary streams used to spread the signal can be generated o in a number of ways The use of shift registers is a common method of generating these streams 8636 1 4 U Such a policy statement was drafted and presented to DoD It was never adop ed The prevailing opinion is that our presentation to DoD was handled awkwardly This failure to grease the skids reduced severely the chances of the policy being accepted Some people however believe that no matter how well it had been presen ted prospects for adoption of the policy were slim The military services are certain to view such a policy as an attempt to invade their domains and they are zealous and effective in protecting what they deem to be their turf Frequency hoppers as the name suggests hop from frequency to frequency within a prescribed set Hopping rates vary from a few hops per minute to many o thousands per second The hopper usually dwells on a frequency for a fixed time but some hoppers with variable dwell times are being produced A binary stream generated by methods similar to those used for direct sequence is used to select the order of frequency use A specific hopper for example might use a table of 256 2 8 frequencies It would employ some device perhaps one or more shift registers to generate a binary stream The stream generated by this device would then be broken into 8-bit segments and these segments would be used as pointers into the frequency table A wide variety of frequency hop radios is being produced mostly by companies in the U S Western Europe and Japan such activities increase so will the likelihood of serious compromise December 1986 CRYPTOLOG page 4 SECRI'3't' r Of I Et A SA BLE fO eO'HRAefORS IIA mLE c P L 86-36 VIA GOMH-IT GIIAP'IP'IELS OP'fLJPYO 1 4 c DOCID P L 86-36 EO 1 4 c 4033692 SEeRE options For example pursuing Option 2 would not preclude some of the research from being done at NSA If Option 3 were ch'osen some of the research and development might be contracted by NSA to industry or the academic community Option 1 Case-by-Case Problem Management U This is the option we have pursued so far When incidents of outside research and development in this area are discovered the particular circumstances of the case are evaluated and an appropriate course of action is selected Advantages D So far we have been able to resolve all the cases brought to our attention in a way that is reasonably satisfactory to all concerned D The services are rightfully jealous of their prerogatives While they might see actions taken under this option as intruding on their turf they might find them more acceptable than the systematic intrusion that would be introduced under Options 2 and 3 CD As a first step in implementing this option DoD would have to require that all government-sponsored research and development on spread spectrum be coordinated with NSA A monitoring system based upon voluntary compliance woul havepi l3'6 of succeeding EO 1 4 c Advantages Disadvantages D This option depends upon controlling by persuasion that which is discovered by chance It provides no systematic way to find out what research is being initiated and despite our success so far there is no assurance that we will be able to influence that which we do discover December 1986 D While the monitoring mechanisms required by this option would make this option more expensive than Option 1 itwould be cheaper CRYPTOLOG page 5 SI enEJqJ PJOT RELEASABLE TO EOPJTRAETORS HANDLE VIA GOMUlT GIIA m i bg O 'lbY P L 86-36 EO 1 4 c P L 86-36 4E3316 J2 SfoJOREJ'f than Option 3 It would also require fewer people with critical skills than Option 3 Disadvantages U Little could be done to implement this option until a DoD policy requiring coordination with NSA on spread spectrum projects was adopted The risk that we could not obtain such a policy is high Even if we succeed a great deal of time is likely to pass before the policy is adopted and promulgated Changing DoD policy is a slow tedious business I d The services would view implementation of this option as an intrusion into matters that have been and in their view ought to continue to be within their jurisdiction They could be expected to resist both the regulation and any control mechanisms we established It would be difficult to write a proposed DoD regulation so that the bounds of our authority were clearly defined The services would likely seize upon any ambiguity in it to justify not coordinating with us or disregarding our counsel U ---tet- COMSEC regulations policies procedures and programs might be used as a model for a FREQUENCY HOPPING FIXED DWELL TIMES E Q u E N C y TIME P L 86-36 EO 1 4 c December 1986 o CRYPTOLOG o page 6 SS6RI 3'iJ 1110T n A A L TO Oll1n A TOf'l IIAIU E VIA eBMfH'f' eHAI4'I4'EL BI4'L f L 86-36 1 1-0 3 3 6 9 2 SE6RB'f' system to control the deployment accounting use and maintenance of spread spectrum equipment containing sensitive components Advantages D NSA's active role under Option 3 would provide much greater control over the technology seepage problem than would the Agency's passive participation under the other two options U Under either of the other options it would be difficult to control duplication of research and development by the individu l services and by us By accepting responsibility for the sensitive components we would not only ensure that the research and development for them was done securely but also that it was done without expensive duplication U Protecting sensitive techniques during the research and development stage is just the beginning Safeguards must be applied to all facets of the product's life cycle Care must be exercised in hoW the system is deployed accounted used and maintained Under Options 1 and 2 it would be difficult to persuade the services to institute life-cycle control procedures that would provide a sufficient level of protection for sensitive components Those life cycle support control proeedures that were employed probably would P L 86-36 EO 1 4 c December 1986 CRYPTOLOG page 7 SECRB'F PdOT RELEA5P BLE TO CmHRAETOR5 HA l fDLE VIA GOMUt' ' GIIAIHffi3L3 61q'L t' aCID P L 86-36 EO LA c 4033692 SEeRS U Openness would not mean disclosing fully everything that we are capable of doing It would however require us to explain to the services what we can do what we can't do and that there may be some things that we can do but cannot provide to them We would also have to provide an acceptable rationale for not sharing some capabilities with them We believe that our decisions on what to release will be based upon sound technical judgments We should be able to convince the services that they will not be made a b i15' GBPi y g 6- uJ of contempt for theIr capabIhttx ' 1 4 c U There are no easy solutions to this problem Option 3 if implemented well would minimize the risk of disclosure of sensitive cryptanalytic techniques at all stages from initial research to field operation of a system P L 86-36 EO 1 4 c P L 86-36 EO 1 4 c December 1986 CRYPTOLOG page 8 SEenEy TO EOPHRAETORS r OT ELEASABLE IIANDbH VIA 60MUf'J' 6IIAHP'fEbS OnLY 4033692 CRYSCO-87 FOURTH ANNUAL CONFERENCE 8-12 June 1987 at NSA Sessions held in the Friedman Auditorium and in Conference Room 2W087 AUDITORIUM SESSIONS ARE OPEN to persons with a green or orange badge WHO IS SHE OTHER SESSIONS REQUIRE TICKETS distributed through your office a green or orange badge and LACONIC access I L - _ _ T54 She was born Hedwig Eva Maria Kiesler in Vienna Austria the only child of a prominent banker In 1938 alarmed at the growing threat of Hitler and the Nazis she left her native land and came to America where she became one of Hollywood's most glamorous stars Her fans never suspected that she was also playing another role that of inventor In 1942 composer George Antheil and H K Markey her married name were granted a patent on their Secret Communication System a way of guiding torpedoes to their targets by frequencyhopping spread-spectrum radio transmissions A technique nearly identical to theirs was later used operationally by the U S Navy General topics include o o CRYSCOs in review IDAJCRD and IDAJSRC updates UNIX conversions Transition to UNIX o Future workstations o Mass storage requirements and proposed solutions Supercomputing trends o Programming multiple CPU and massively parallel systems o Requirements for software management system o Computer graphics o IOSP updates o Software development standards do they work Who is this actress-inventor Answer on page 20 o Reports from CSE DSD and GCHQ o OWL libraries o CRYSCO-87 wrap-up A5381P13 963-4196 P L December 1986 CRYPTOLOG FOil OI FlCIAL U E page OP-tLJPY 9 86-36 DOCID 4033692 QUOTE WITHOUT and is more or less a fort equipped with the most advanced electronic warning systems 11 The staff are chosen from among the personnel of other security agencies after being COMMENT subjected to psychological training courses and a Editor'sNote A number of readers sent us copies of detailed scrutiny of their personal files Because the following translated excerpts from an article wives love to gossip only bachelors are allowed to that appeared on 25 March 1986 in a literary join the NSA The staff may only marry each other magazine ofthe United Emirates Republic 11 The staff is only allowed to have the minimum amount of social relations or may not have any social life at all 11 Only a few people know that this agency actually exists almost no one outside the US knows anything about it while one out of a hundred thousand Americans knows of its existence 11 11 The NSA has very advanced computers one for analyzing normal calls and the other to pick up and analyze coded calls The average number of calls that the computers receive daily is 300 000 normal calls and 200 000 enciphered calls These computers are able to read and to analyze codes at up to 600 lines per minute and in all languages 11 11 The agency is a field study' of the most upto-date scientific advances 11 It possesses almost absolute power its ears and eyes are planted everywhere but mainly in the Soviet Union NSA personnel were the first to know of the crash of the Russian space shuttle Soyuz 11 Like the human brain the agency's brain' is divided into two parts The right part named CARlON ' is equipped with four linked IBM 3033 computers attached to three huge printers that are able to print 22 thousand lines a minute The left part is equipped with super computers each of which weighs 5 000 kilograms and are able to do 200 million operations per second NSA computers write up to 320 million words per second 11 11 The annual budget of the NSA is $12 billion and it employs sixty thousand Americans Its chief assignment is to know everything not only inside the US but everywhere in the world It traces all telephone and radio calls whether their subjects are political military economic cultural or even personal Its equipment and staff are able to analyze ciphers Outside the USA the agency has secret operations branches in Japan Taiwan South Korea West Germany South Africa Lebanon and 11 It is not an exaggeration if we say that if Turkey These offices are usually described as Gorbachev sneezes the Whi te House will be the annex buildings to American embassies first to hear it consulates and US cultural and commercial 11 The agency was established in 1952 during organizations the second term of former president Truman It is 11 The agency relies mainly on satellites that not subject to the oversight of the congress or any are specialized in espionage From an altitude of other US organization Furthermore it does not 200 kilometers these satellites are able to legally exist When it was established only the photograph any object that is two feet or more president and national security advisor knew above the surface of the ground about it 11 The reason some of the NSA's secrets are 11 The NSA headquarters is located inside a being revealed at this time is to preserve part of thick forest of cedar trees near Fort Meade America's image of technological superiority which Maryland two hundred kilometers from was injured after the CHALLENGER tragedy 0 Washington The complex is very well protected December 1987 o CRYPTOLOG o page 10 FOR OFFICIAl- lJSl'J O bJPY DOCID 4033692 SHGRKT The Language Identification Problem u ------ P16 P L 1EURt- A -message in a language that no one can recognize is a vexatious thing there is no - knowing whether the information is trivial or important as long as the text cannot be underStood Even if the information in itself is not valuable there is the possibility that identifying the language may be helpful for _ making decisions and adjustments in collection 000 The first thing we have to do is to get rid of the idea that language identification is simple Language identification is a large subject that has only begun to be explored along lines that we would find useful To begin with consider the problems I The way in which language recognition takes place is not well understood It appears that all people do not do it the same way some rely principally on recognizing individual words and others primarily on recognizing rhythm cadence and overall effect I Estimates of the number of languages in the world have run mostly between 3000 and 4000 and have tended to keep rising The number of languages in Africa is said to be at least 850 in India 800 in the USSR 130 in South America with less certainty perhaps 500 or more and so on 86-36 I In the future we can expect to get more intercepts in unidentified languages than before This will be due to increasing international travel and commerce the use of foreign labor the gr9'ying-availability of radiotelepho g e satE llite communications and multiple channels and new collection techniques with more comprehensive intake POUO Fortunately for us while 4000 is a meaningful figure in the field of anthropology in the world of communications it is not In practice we are not going to have to identify 4000 languages The vast majority are minor languages spoken in remote and primitive areas and by tiny populations some of them in single villages and 30-person tribes U Of the 4000 languages about 100 are spoken by 3 million people or more and another 50 or so are spoken by 1 or 2 million all the rest have fewer than a million speakers apiece Only about 70 languages are official languages of a national government Thls is because Arabic English French and Spanish are used by so many The number of countries in the world is about 170 Another 15 or so are languages of sizable and well-known minorities iEURr In real life we are going to find ourselves dealing with the same languages most of the time and there will not be much more than December 1986 CRYPTOLOG page 11 SBOREltf HAUDLE JPYfA OOMU 't' OHANNELS mny DOCID 4033692 S CftE't' 100 of them which is still enough to give Look for a short speech segment of one or two everybody plenty to do My own experiences syllables or words have generally fitted in with this principle U Often the easiest way to identify the When it is an official national language we can usually identify it when it is a well known language of a conversation by word recognition minority language we have a good chance and is to concentrate on the shortest speeches in the exchange Short speeches by nature present when it is neither of these it can become very the fewest word division problems tend hard going strongly to stereotyping and the use of common words and usually represent simple ideas '- fflt- How do we deal with all this Experience has brought some pieces of accumulated wisdom with respect to both voice and graphic communications as well as a subset of the latter handwriting which must be considered separately 4G - The ideas I present below are based on my own experiences with language identification These'ideas are set forth as suggestions for linguists and ways that managers and supervisors can help U The three primary short speeches are Yes No and What A highly favorable case occurs when there is a long speech of one or more sentences a short response of only a syllable or two and then a repetition of the long speech The odds are that the- short speech means What Or if the response is longer than one or two syllables it is likely iQ---be Hows that What did you say ''Wh'iit was that ' didn't get that or I co'uldn't hear you U Now you will find it much easier to check for a word or phrase in dictionaries or lists when you know what you are looking for Unfortunately there are people everywhere in th-e-world who say Aahh or Haahh instead of What Keep an Open Mind and Try Again I POINTERS FOR LINGUISTS F'OUO Following are some suggestions for linguists who have been tasked with identifying an unknown language Voice communications message traffic often referred to as graphic and handwritten material are each considered separately A VOICE COMMUNICATIONS Try Rewind F'OUO Probably the first thing to do with a voice tape that defies identification is to try running it the other way in case it has not been rewound Don't laugh a remarkable number of unidentified languages have turned out to be Russian or English in reverse EO 1 4 c P L 86-36 December 1986 IIA TB bE Consider the possibility of intrusive foreign words U International word borrowings contaminate language samples and have a cluttering effect that makes identification harder Words of international scientific and CRYPTOLOG S Cft 't' page 12 VI A OOMHH OIIJd'Hf8LS OPfLY EO 1 4 c P L 86-36 EQ 1 4 c DOCl D 4 03 S 6 9 2 6-36 Establish What is Good Text by Findin g Repeats U Make a photocopy of the text so that you can mark it Then underline every string of letters that occurs in the text more than once This will usually ensure$at the things you are looking for in dictio GBP s ret garbles EO 1 4 c Collect References technological vocabulary like telephone are so widely adopted that they only hinder identification U Languages of Islamic countries have a considerable stock of borrowed Arabic words and frequently it is the same Arabic words that several unrelated languages have borrowed Obvious examples are Persian Indo-european Uzbek Turkic and Indonesian Austronesian which have many of the same Arabic words in common U There are several identification books intended for librarians and printers which devote a page or two to each language and give a paragraph of sample text Some of them also give some information about the grammar and point out some common words that are characteristic One drawback is that the sample paragraphs are quite short and often fail to include some very characteristic words Another is that the samples are given as they would appear in a printed book whereas traffic is often in transliterated form and there can be several transliteration systetns for the same language C HANDWRITTEN MATERIAL FOUO Handwritten materials present special problems and must be considered separately Generally they present the hardest problems on 86-36 EO 1 4 c P L December 1986 o CRYPTOLOG o page 13 SEJ6REJ'P IIAoPWLEJ JPYEA GOMI lT GHA lnK bS O r bJPY P L 86-36 SEJEURRS'f the In'anhic side I language either in plaintext or in simple substitution After that we are in for a long effort with poor prospects Handwriting that can be matched to a known cursive script such as Arabic Russian Greek Armenian Georgian or old Albanian is easily identified Samples of cursive writing can be found in the following books Call numbers are those of the NSA library U Diringer D The Alphabet 1968 P211 D62 1968 2 vols POUO An example of all of these problems is the Voynich manuscript a unique European manuscript thought to date most probably from the 15th or 16th century which has resisted solution not only by philologists early in this century but by NSA cryptanalysts as well Jensen H Sign Symbol and Script 1969 P211 J45gE 1969 Hoskins Meredith-Owens A Handbook of Asian Scripts P213 B77 U What is not easily identified is handwriting with partial resemblances to one or more known scripts especially when the characters are connected to each other creating great uncertainties about how to separate them Handwritten pieces of this kind present too many questions all at once Is it plain language in an unidentified language Is it really writing at all or just doodling Or done by someone under the influence of a narcotic who thought he was writing Or a hoax If it is plain language why can't we match it with a known script Is it a cipher If so we are at a great disadvantage in not knowing what the underlying language is and at another great disadvantage in not knowing how many characters there are and what they are U For example if we were looking at an unknown script that contained signs like English handwritten c and cc e and ee I and 11 t and tt all letters made with only one or two strokes we could not be sure whether cc ee 11 and tt were connected double letters or separate letters in themselves U Again we can start by looking for repeated sequences which would suggest that the text was not random doodling but would not prove that it was not and for repeated combinations followed by spaces which would suggest grammatical endings in an inflected December 1986 F'OUO Among the many problems are the alphabets sometimes bizarre that are invented by the writers fortunately these often prove to be monoalphabetic substitutions of a wellknown language and so they are easily solved An important diagnostic feature of invented alphabets is that the characters are usually clearly drawn and clearly separated II HOW MANAGERS CAN HELP Simply having a list of expert voice linguists who can be consulted has been tried it was found that the list required constant updating because of normal employee turnover In any case the old practice of carrying a tape around and trying it on various linguists is slow inconvenient unreliable and not always successful either It is also heavily dependent on luck and guesswork in finding the right linguists to ask U But there are several other ideas that might be tried to improve our success rate in identifying unknown written text And these depend upon the active support of managers and supervisors Messages in unidentified languages tum up only sporadically but persistently Because they constitute a recurrent problem but not a continuous one they are dealt with individually without any method or system and without any record of what was done and what resulted They should be dealt with more methodically What I have in mind would not involve concentrating and centralizing identification efforts but would be the opposite of that making more analysts able to do identification by providing information on it CRYPTOLOG page 14 SHCRE't' IIldfDLB VIA GOMIH'J' GIL l'ar aJg Ol'Ib Y DOCID 4033692 SElCREltt' THE VOYNICH MANUSCRIPT u ' ' -' ' December 1986 CRYPTOLOG page 15 SIWR T MA10fBLE VIA COMIU'f CHANNELS ONLY ' DOCID 403369 L 86-36 SHEURRH'f' more systematically By no means are all of these new ideas Support the Updating and Expansion of Existing Word Lists D In general supervisors and managers can do two things to help linguists identify unknown languages Support the Development of a Training course FOUO Training is not a new idea It would have the advantages of introducing methods in a package rather than have everyone rediscover them individually of furnishing supervised familiarization and practice and of providing a basis for the creation and distribution of working aids For example the course could include a brief and practical introductic n to phonetics with special attention to speech sounds in a wide variety of languages Commission the Compilation of a Word Grid Provide Reference materials U Recognition is by nature matching The analyst must be provided with bases for comparison This may mean purchasing books subscribing to periodicals and acquiring video and other voice tapes U Now for some specific suggestions about each type of language identifcation problem FOR VOICE TRAFFIC Encourage the Use of the Tape Library P L FOR A WORD GRID Yes No What Hello Hi Goodbye So long Thank you 86-36 EO 1 4 c December 1986 P L 86-36 Wait a minute Who is calling All right Okay That's ri ght I understand I see Listen Don't hang up That doesn't make any difference Fine Excellent Impossible Immediately Necessary Important Will arrive Because North East South West Yesterday Today Tomorrow Tonight Day Night Week Month Year Next Last Week etc Names of weekdays Names of months Numbers from 1 to 10 Selected higher numbers e g 100 1000 CRYPTOLOG page 16 SH6RE'J' IIAUDLB ViA OOMHi' ' OIIANPH3LS ONLY DOCID 4033692 SRCRR'f' amount of text and without too many borrowed or foreign words Collect Samples of Transliterations and Romanizations 8-000 Printed books on languages regularly give each language in its traditional writing system I Sponsor In-house Research CD Possibly some of the many things that are not known in language identification could be resolved by study and investigation For example there appears to be a difTerenceof opinion on whether it is possible for someone who doesn't know a tonal language to tell whether it is one merely by listening to it The objection is that in at least some cases it may be difficult to tell word-intonation from sentence-intonation If this question_cQuld be settled it would be a valuable piece of information in language identification P L 86-36 A FINAL WORD D There is a great deal to be done in language recognition At this time no one knows how many possible approaches there are and no one system can claim to be invariably successful D December 1986 CRYPTOLOG page 17 SEleRE'P IIAPWL13 YIA OOPtHn'f' OIIAPH 13LS OnLY DOCID 4033692 CONFIOfi3r 'fIAL PRODUCT vs PROCESS One View u L 86-36 ---_-- o------IG14 FOtJO Like the management of many other middle-aged organizations the management of NSA is forgetting why the organization was created The auxiliary functions of the Agency are distracting us from our primary task of producing intelligence information In other words we have become enamored with the process at the expense of the product Recognition of this situation is critical because there is no bottom line to measure NSA's effectiveness there is no danger signal to alert us that all is not well Some may argue that customer satisfaction is the bottom line but I believe that the customer is not aware of what could be provided and therefore makes decisions regarding NSA product without knowledge of our full capabilities 'I I December 1986 o CRYPTOLOG o page 18 CONFlBl P 'FlAL P L 86-36 EO 1 4 c DOCID 4033692 eOP'JFIBI3N'-FIAIJ o P L 86-36 EO 1 4 c December 1986 CRYPTOLOG page 19 eo PH'll lS Pf-PIA L DOCID 4033692 P L 86-36 EO 1 4 c Answer to WHO IS SHE on oaae 10 0l AlleJnleu sawo s6u 4l 'P l S a uo 04M 'JJl Wel ApaH i'H lJ 6U IIOJdwl December 1986 CRYPTOLOG page 20 eOpt l'lDI3N'HAL IIA fDLE VIA GOMUt' ' GHAK fELS O fLY DOCID 4033692 r'-'-'-'---'-'-'---'-'-'---'-'---'-'-'-'-'-'-'-'-'-'-'_ - _ _ p L 86 C36 LW g221 _ _ _ _ _ _ - _ _ - _ _ _ _ _ _ - - _ - _ _ - _ _ _ _ - _ _ So you want to change jobs within the Agency One of the great advantages of working at NSA is the vast array of varied opportunities awaiting you Unfortunately many employees find it difficult to move It doesn't have to be The first step is to determine your goals For some the goal is to become a senior executive while for others it is the more modest goals of enjoying their jobs and receiving reasonable promotions Some employees want to work within a particular discipline and move around among different targets and others are enamored of a particular country but want to work on various aspects of that target still others want to diversify and seek assignments in other fields such as personnel logistics COMSEC etc And then of course there is the question of management or tech track Once you have examined your December 1987 personal needs and determined your goals you can start looking In order to be able to move easily you must first establish yourself as a skilled and willing worker Get your assigned tasks completed and then ask your co-workers to teach you what they are doing Ask your supervisor to give you tasks outside your normal range of duties Learn everything you can Take every course in your discipline that you can arrange There are self-study courses in abundance and you may find a learning center open after or before your normal duty hours Try changing jobs within your current organization that's generally an easy move Once you have established yourself as a skilled experienced and willing worker start the above process in reverse Teach your skills to others in your office train someone to do CRYPTOLOG page FOR OFFICIAL US ONLY 21 DOCID 4033692 your job or train several people to do different parts of your job The mission your coworkers and your supervisor all benefit from this approach and so do you This course of action will allow you to move for it is often easier for managers to find and hire someone at a lower grade Don't fall into the trap of trying to make yourself indispensable if you succeed you are doomed to stay in that job Write a complete and detailed description of your job No one is more capable of doing this than the individual doing the job This has multiple benefits Not only will you surprise yourself pleasantly in finding how complex your job is but by detailing it you may think of some ways to broaden or simplify the tasks or to improve on your performance The description will also serve as a detailed guide for your replacement for learning your duties In addition your supervisor will be able to use this job description when writing your performance appraisal as well as a guide to writing a vacancy announcement and even as a tool for interviewing applicants for your job Use your network of friends and acquaintances to look for a possible successor Even if your job is all wrong for you you may know someone else who would be a perfect match for your job Suggest they come and talk with your boss In fact everytime you hear of someone looking for a job change or hear of a job opening try to help All this and we still haven't gotten around to discussing how you are going to find a new job Well being in a position to be granted a release from your present job is the December 1987 first step in moving Without it you will have to find an advertised job and even then your supervisor can hold you for 90 days Obviously one of the secrets to moving is making it easy for your supervisor to let you go OK now you are ready to move How do you find that new job My first suggestion is to review your goals looking for various intermediate paths to achieving your goal You can stay in one place a long time waiting for that one perfect job or you can be flexible and take a series of jobs all of which teach you new things and allow you to make various contributions Now start talking to the people in your network Of course you have one They are the folks you eat lunch with members of any Agency clubs you belong to people you deal with professionally in other offices If you've been trying to help them they'll help you Find offices which interest you and make an appointment to talk to a branch or division chief in that office Ask your boss to check out his or her network for openings Go to your Pers Rep and your Career Panel Exec and let them know you are looking Don't assume that all advertised jobs are bummers Talk to anyone who has a job which might in anyway move you closer to your goal The job might not have the right title but the duties may be exactly what you are looking for or close enough to fill the bill Remember that you can apply for a job one grade above or below your current grade No you don't take a salary cut Get your Personnel Summary in tip top shape and have copies available Make sure you have neither under nor over- sold yourself Find ciritical friends to review it for you Ask former supervisors if you can use them as references If you are a GG-12 through GG-15 make sure that the SPCP folks M44 are aware you want to move Be creative innovative even daring in looking for ways to promote your job search A large portion of Agency jobs are constantly being vacated and the opportunities are there waiting for you CRYPTOLOG page 22 FOR OFrI6IAL USE OP iblf DOCID 4033692 eop PIBE N't'IAL Having found some interesting openings interviewing for the position is the most important thing you can do Many even most NSA supervisors have never been trained to interview job applicants This means that you must train yourself to be a top notch interviewee without having the benefit of being asked the right questions You have two goals during the interview finding out if the job is what you want to do and finding out if you and your prospective boss share the same philosophy Typically an Agency supervisor will tell you what is being done in the office and a general idea of what your job will be You will have to ask specific questions about the actual work In some cases you will want to spend some time with the incumbent or another worker to see what is actually being done This is the easy part If you want to be happy in your new job you also have to reveal a lot of yourself to your prospective supervisor Tell him or her your worst faults--if he or she can't live with them neither of you will be happy when those faults come out on the job You're better off knowing that up front Also discuss what you believe to be your virtues and your methods of operation If you need a lot of independence on the job say so you'll chafe under a controlling supervisor If you need regular feedback on your performance say so you'll get ulcers waiting for the strong silent type Watch and listen carefully to the supervisor during this discussion Body language and verbal responses to your revelations will tell you more about that manager's style of management than you could ever learn by asking What's your management philosophy Be truthful about your intentions during an interview If you plan to stay in the new job for only two or three years make that plain You can contribute a lot during that time and anyway there is no stigma to moving the seniors do so about every two years If at the end of your interview you know that you aren't interested in that job have the courtesy to tell the interviewer clearly that you are not interested You might even suggest the name of someone you know who might be a good candidate for the job If you are considering several other jobs you should tell the interviewer this fact Play it straight this same manager may call you years later with a job you do want 0 The following is classified CONRBEr'vftAL in its entirety B @ @ @ ODE TO THE CA INTERN PROGRAM u 0 fill 86-36 53 IB J ffi1 fQ Intern Grads both brave and tough How we learned that cryptic stuff raJ Was that program quite a chore And that paper quite a bore @J @J Did we importune our buddies As we practiced Folklore studies 0 fi ffi rliI fi J Did we grow more fatalistic As we mastered each statistic fill If we shunned some probability Did we suffer culpability raJ In the washroom did we primp Rather than practice IMP I' J Would we games and puzzles play Instead of work upon a CRA Y Or struggle with a CDC Or worse yet some weird SPD IB raJ raJ To choose UNIX or PCDOS Were we ever at a loss @J @J @j @J @J @J @j @j @j @J 0 @J @J Feezy filzy foozy fumzy Do our brains feel slightly numbzy @J @j @J Ii Are we live or are we dead After so much traffic read @J @ rEl fID Now we're Permanent and Real What's the bravest thing we feel DO YOUR WORK OH CAFETERIA AFTER THREE YEARS f' I WE'VE NO FEAR 0' YA @J @J @J @J @j lQ _ @J lQ @@@@@@@@@@@@@@@@@@@@ Repnnted WIth permission from the October 1986 issue of the Cryptanalysis Intern Bulletin December 1987 CRYPTOLOG o page 23 eOPH'IlJAPHIAL DOCID 4033692 Loosely Q3 Q CIJ CD Q o s lY b -- CD CD I n CD P L 86-36 1 laid ooy 1 sure could use SOille computer suplJort I've got masses of data in this langua6e to process and I need some efficient way to put ttle 600d stuff into english de laid ooy are you in luck I've got just the thing for you an interactive virtual UIltra-fast Gizmo Super-X computer with a superpowered glitch-free VUT that won't even give you eyestrain It will do everything for you but tie your stlOes 1 laid nut I don't need my shoes tied I need help translating your macnine do that Can He said Of course It can interact instantaneously in til different scripts including upside-down Urdu complete with all those funny squiggles and dots whatever they are I never claimed to be a linguist myselt 1 saUl but I do this laIlC uage not Medieval 110ngol and my output is supposed to be all in english I don't compose in all these other writing systems Ie said 1'0 proolem we'll Just ignore these ott er scripts they only cos t a few ttlOusand dollars per luachine anyway do you need December 1987 o CRYPTOLOG o page 24 POR OPPI6Iz1t L USB Ot LY l'OW what else DOCID 4033692 I said Well how about a really sizeable dictionary and the capaoility to look up a term more quickly than I can by hand He said Of course you know that will taKe all your memory But we can do it If you'll j st fund the typists we can probably get the dictionary entered in a couple of years And the neat part is that you can program'the dictionary yourself I said 'le progr m I'm a translator I don't have time to learn to do all tilat - -- Not to worry it doesn -t take all that long I picked it up in just a few weeks myself Just think you'll be able to do all your own programming Only thing is the course is full and you can't let in until sometime late next year ' -l sald Why can't you do the programming for me said WhO me I'm too busy for that I already spend nearly full time as it is working the bugs out of the programs other people write I don't need another job You know these systems anyway they're all overloaded Overloaded already Then how can I add a program wi th dictionary look-up capability word proces sing and all the r est Well you really can't for now But just as soon as we get the Gizmo Super-X you--Clln do everything since it 'will stand alone sa id You mean it won't have to talk to all these other systems ilut what if I need to do just that Well it won't unless we develop a special package but I'm sure that's already well in hand Not to worry So I have to wait for the Gizmo Super-X plus my programming class plus whatever interface package eventually gets developed all before I can start work With the budget situati on the way it is these days What do I do in the meantime well if you only knew how to pr0 lram you could use your present system for oh just lots of things of course it's overloaded so don't try to add too much maybe just a few working aids or so And by the way the word-processing program on that system takes too much system space so we're going to take that out Thanks a lot said tlave another pencil ' I------------December 1987 CRYPTOLOG page FOR OFFICIAL USE 6NLY 25 DOCID 4033692 eONF'IBBN AL This review is classified eOnFIDEf'ff'IAL eeo in its entirety GCHQ The Secret Wireless War 1900-1986 By Nigel West 294 pp Weidenfeld and Nicholson London 1986 LI2 95J Reviewed by Vera Filby E41 This book on NSA's senior partner is a w'eUorganized well-researched history of British '-- SIGINT beginning with its origins in telegraphy and wireless and carrying through in straightforward chronological order to events current at time of publication in 1986 It is no expose or Bamford-style effusion though the author has a reputation for such writing but a balanced account of what happened how it happened and what it means with abundant detail on organizational developments complexities and interrelationships on codes ciphers and cryptomachines on cryptologic places and events all over the world and on people Readers can enjoy the opportunity of learning more about some or-the extraordinary people who over the years have made cryptology and SIGlNT what they are BOOK Il EVIEW This is also a book for entertainment as any SIGINT history can be expected to be since SIGINT in all its intricacy is endlessly interesting and a continuing source of great stories -- stories of achievement of loss and failure of discovery stories of rivalry international intrigue war and betrayal What a goldmine historians have found to dig in since the ULTRA revelations in 1974 Some like Nigel West seem to have developed an understanding and appreciation of SIGINT's special nature West is a military historian and this is his seventh book on intelligence matters Except for some private records of the Radio Security Service his listed sources are public records and published works L --- - -- - -_ J ntil the beginning of the first World War British authorities preferred to trust the security of their cable links throughout the Empire and were wary of wireless because of its vulnerability to intercept and its potential for spying When war started radio amateurs were banned but eventually they were called upon to help Future cryptologic organization was foreshadowed in a recommendation to make a list of those who would be willing to be trained in encryption decryption censorship and interception What was to become SIGINT started when wireless experimenters in the field in France came across German telephone conversations That was the first surprise The next was the clear evidence in those conversations that the Germans were intercepting the communications of senior Allied officers War Office intelligence could hardly miss the significance of-this and soon the first military SIGINT unit was founded Thereafter and throughout the war SIGINT became an increasingly indispensable asset for the military forces The cryptologic history of the Great War is well documented and its outstanding events most notably the still controverial battIe of December 1986 CRYPTOLOG page 26 eo P I' I P 'I'IA L HAnDLE VIA eOMHf'f' eHAf'ff'fELS Of' LY DOCID 4033692 eONFIDElN'f'IAL Jutland and the celebrated Zimmermann telegram -- often recounted West retells this history as an integral part of his total picture In an effort to reduce extensive service duplication a new body with the innocuous name Government Code and Cypher School was established on 1 November 1919 With wartime targets gone GC CS redirected its efforts to Soviet traffic which was supplied by Army signals elements and to Japanese copied by Admiralty stations In addition civil and governmental communications provided a rich supply of Soviet diplomatic traffic much of which was readable and highly valued for its abundant evidence of Comintern programs of subversion In 1920 top British Government officials chose to release decrypts of Soviet telegrams with all surrounding details to expose Soviet undercover activities No reaction was apparent in Soviet traffic but then the Government did it again By the end of 1920 the Soviets were using couriers and the traffic had disappeared When it later resumed it was quickly broken and was read until May 1923 when once again the Government blew it this time to challenge Soviet adherence to a Trade Agreement of 1921 and to support an ultimatum demanding that the Soviets cease financing subversion in Great Britain and the Empire The price of this political decision was again temporary loss of the traffic Finally after yet another deliberate compromise in 1927 the Soviets introduced one time pad These are the things that break SIGINTers' hearts In 1922 the Foreign Office took over administration of GC CS because it was believed that the central problem for SIGINT would be diplomatic and it was given jurisdiction over intercept DF TA and CA in order to centralize control over these scattered functions At that time the staff numbered 91 Led by a few brilliant survivors of Room 40 the organization achieved many successes but remained in a state of underfunded obscurity up to the very edge of war Only in 1938 was a German section formed But by the end of 1939 GC CS had acquired the wartime covername of Government Communications Headquarters and its staff had grown to 937 The wartime peak in June 1944 was 6 812 The Enigma machine enters the story when in 1926 the German Navy began using a version of it The British Admiralty had bought two machines in 1928 and after years of consideration a Whitehall committee decided to have the Air Force supervise construction of a machine based on it This became the TypeX which provided secure telecommunications for Allied SlGlNT throughout the war Experts recognized that Enigma machines in all their versions were impenetrable if used correctly But with thousands of German Enigmas in use and under all kinds of conditions of stress and crisis errors did occur and these compromises did help the cryptanalysts break into systems The fact that error can make systems vulnerable is well understood and it is therefore rather surprising that the author makes so much of it It is his view that this truth is the real ULTRA secret istorians seem to have been so intrigued with codes and ciphers and so involved in reevaluating battles campaigns and diplomatic maneuverings in the revealed context of intelligence that they have been less concerned with researching the fundamental role of intercept At the time of his death Ronald Lewin author of ULTRA Goes to War and The American MAGIC was compiling materials now deposited in Churchill College Cambridge for a study of the British military intercept services which have been coordinated by the Y Committee since 1928 and are referred to as the Y Service The Story of the Y Service was the subtitle of Aileen Clayton's excellent history and great adventure tale of her wartime RAF service The Enemy is Listening published in 1980 West has now added to this record his account of the Radio Security Service True to the tradition started in World War I wireless amateurs volunteered to help the war effort Their help was sorely needed because December 1986 o CRYPTOLOG o page 27 CONTmEN'-FIAb HANDLE VI c COMHfT CIV't r fFJLS O fLJPY DOCID 4033692 CONFIf NrrIAL Foreign Office and Special Service facilities were ineffective or inoperative The role of these Volunteer Interceptors when they were organized into the Radio Security Service in 1939 was to cover agent communications By March 1940 they had completed their mission the target had been eliminated Meanwhile they had been copying great quantities of enemy signals from the Continent developing techniques and building data bases Of their contribution West says that it was largely due to them that the cryptographers at Bletchley Park were able to continue their work A history of GCHQ necessarily includes a history of NSA and West begins his treatment with the closure of the Black Chamber in 1929 and continues it as part of a combined story up to the present Negotiations between Britain and the United States began in 1940 and the next year a replica of the PURPLE machine which produced decrypts designated MAGIC of Japanese diplomatic messages was sent to GCHQ headquarters at Bletchley Park also known as BP Station X and to the Navy HMS Pembroke PURPLE had by then survived an appalling leak when American authorities disclosed to the Soviet ambassador information in a MAGIC decrypt concerning Hitler's intention to invade Russia But those who as it turned out could have profited most from MAGIC did not have access to it In his brief summation of Pearl Harbor West notes that the commanders were out of the loop and that the low level tactical traffic that would have warned them was processed after delay in Washington where the strained resources were committed to traffic of presumed higher value UK and US SIGINT collaboration was formalized in the BRUSA Agreement of 17 May 1943 This and later agreements with Australia and Canada established the basis of the structure that exists today So productive was the collaboration that by the end of the war the SIGINT effort was flooding intelligence channels with torrents of information clever author likes to surprise his readers every now and then and West does so by switching suddenly from the SIGINT scene at the end of the war to Australia In 1944 a station there -- of all places -- picked up Soviet traffic in a system codenamed VENONA This event West believes helped set the stage for a peacetime mission in exploitation of Soviet clandestine traffic This traffic revealed continued Soviet involvement in and support of international subversion -- the story of the 1920s unchanged Exploitation of VENONA led eventually to the identification of the spy atomic scientist Klaus Fuchs the traitor Donald Maclean and other spies and traitors in the ramifications that continued for years The last two chapters of the book covering the postwar period portray a scene of continual disasters from the shooting down of the EC-130 over Armenia to Prime and the Pelton case This is not really surprising since only the things that go horribly wrong get out to become meat and drink for historians and investigative reporters But such a recitation of losses leaks compromises defections and multiyear penetrations leaves a SIGINT reader feeling like a farmer who has seen his entire community's long-nurtured crops Wiped out by insects and disease The author manages to wind up however on a positive if cautionary note The secret wireless war will continue for as long as there are communications to be intercepted and signals to be interpreted Accordingly GCHQ will remain an invaluable source ' and a tempting target This book has been faulted for lack of proper scholarly footnoting and original research Be that as it may for readers who want only a consolidated readable survey of the publicly available record it will do the job 0 December 1986 CRYPTOLOG o page 28 eOP l'lnSprrlAb IIAUDbE V A COMln'F CIIA f lEL8 O fLY DOCID 4033692 architecture The SIMULA 67 language was used to investigate the 4th layers of the protocol A 'problem oriented' network 4 SET is a distributed computer system based on the IZOT 1016 small computers and the SM-4 mini computers The SM-4's are connected to one another by data transfer channels and form a network which operates across large distances A term 'teletreatment of data' is introduced in connection with a description of the experimental packet switching network SET 1 2 The possibility of using optical fiber links for high speed transmission to perform 'distributed data treatment' is investigated in one paper TECHNICAL LITERATURE REPORT Six papers deal with LANs Foreign LANs such as ETHERNET IEEE Project 802 ECMA Sinpads and Hypernet are investigated to determine the control functions and access functions No specific Soviet LAN is mentioned An interesting feature of the conference is the variety of different LAN ami packet net schemes presented There does not appear to be a single national system and some of the regional systems link into the AKADEMSET net This shows local variety that seems surprising in the centralized Soviet system '1CM HOH'b A1tz A-UU _ 'p '''' W - ABSTRACTS ON SOVIET PACKET NETWORKS U P L 86-36 I i Reported by ' Ip 13 The June 1984 issue of In spec Key Abstracts on Communication Technology contains translated abstracts of papers presented at the 3rd All Union Conference on Computer Networks for Packet Commutation BblqHCJIHTeJIbHble CeTH KOMMyTaU HH IIaKeToB Te3Hcbl J oKJIa n oB TpeTbeH Bceco103HoH KOH j epeHUHH in Riga Latvia in September 1983 The Russian proceedings were published in two volumes with abstracts in each volume One foreign network the Spanish IBERPAC was described in the translated abstracts reported here One of the papers notes problems in implementing X 25 that result from the lack of a standard formal description of X 25 and X 2l In Bratislava an experimental computer network IPK is being developed using the ISO December 1986 The Soviet Academy of Sciences uses an AKADEMSET packet switched network for internal and foreign information-computing resources The Ministry of Communications and parties with data bases and applied software can also use the AKADEMSET network Elektronika 100-25 minicomputers are interfaced by hardware adapters to the BESM-6 computers and the whole structure used as an interface machine for accessing the AKADEMSET network The SM-4 TISA-4 has also been proposed as an interface with the packet network AKADEMSET X 25 packet transmission protocol is used to connect the ES7920 computers to the packet switching center in an open network There is apparently more than one packet net development and several different services A subsystem EKRAN of the DNEPR computer network is used for electronic mail A network of the TsSu SSR provides user services which include transfer of single- and multipleaddress information virtual communications between two points data collection by computer CRYPTOLOG FOR OI ICIAI page 29 y Q I y DOCID 4033692 from user terminals dialog communications between networks users and operator-user communications The MSSR Academy of Sciences has developed a distributed computer network for quick access by a large number of users to an automated data base of scientific information A multifunctional terminal system of a scientific center uses a central SM-4 minicomputers and peripheral Electronika-60 microcomputers connected through the recommended X 25 protocol This provides software exchange collective use of external system hardware and simulta neous remote interaction among several users based on variable packet lengths less than 1K and data transfer speeds between 50 kbytes s and 1 Mbyte s A Terminal VTsKP network for regional control of internal affairs combined different computers into a single distributed terminal network using hardware developed along the CAMAC standard A regional computer subsystem 'Ural' is intended to give users a wide range of services including remote access to the subsystem informational-computational abilities as well as access to the AKADEMSET netwOlk The DNEPR network has a communication network that will accommodate 255 ports receive or send messages up to 27 Kbytes and remotely open or close ports A synchronous communication controller was developed to do bit-by-bit signal treatment using HDLC high-level data link control SDLC Synchronous Data Link Control and K 25 2 probably a typo for X 25 2 type protocols The interface is intended for a microcomputer and connects with the main bus from one side and with either a synchronous or an asynchronous modem The use of a subsidiary processor to free a main computer from queues of data will prevent inefficient data treatment even though it makes the net more complicated One paper proposes the use of a TISA-4 computer to provide interface for a X-25 is the standard packet mode transmission in more than 50 packet switched data networks that are operating in other countries It is a OTT recommendation for a packet protocol which makes a universal interface feasible between data terminal equipment and public packet switched networks It is defined on three levels viz physical link and packet level The digital communications are defined by ITT Recommendations X 20 and X 21 for stopstart and synchronous communications and most existing packet nets support data rates of 2 4 4 8 and 9 6 kb s Some also offer 48 or 56 kbps The Soviets are currently using modems on some circuits which meet ITT 4 8 kbps specifications One of the Riga papers specifies a LAN storage of systems software and ini tial loading of separate microprocessors in the network Analysis The US began its development of packet networks about eighteen years ago with ARPANET which had heavy university and laboratory participation Since then a number of packet nets have developed many with scientific and resource-sharing aims The Soviet AKADEMSET appears to be a development along similar lines The technical features as far as can be determined from the abstracts are centered around micro and mini computers and medium-speed data modems The work is probably about ten years behind US applications complex of SM-4 and BESM-6 computers with the AKADEMSET network via the network access method in TISA-4 Another proposal would use specialized hardware to perform the lower levels of the packet switching protocols to free the computers for more problem solving AL ------------- - ------- I cylindrical magnetic domain memory perhaps some kind of drum or disk system of 100 Kbytes is proposed to give a reliable storage for EO 1 4 c 86-36 December 1986 o CRYPTOLOG' page 30P L FOR OFI ICIAb US i ONt JPY DOCID 4033692 SECRECP Leete P L 86-36 To the Editor lJ I chuckled to myself the other day as I readl lintrospections into the misty world of collection management I think we all go through this period of doubt and reassessment occasionally - I know I have Whether it is brought on by the myriad of reorganizations we have experienced by endless criticism or by mid-life crises it is probably healthy to step back periodically to recount and reassure ourselves i e Now that you are a professional collection manager what are you expected to do for a living December 1987 IIA mLE YI ll CRYPTOLOG page 31 SECRB COMH'f'f CHll fP tELGBP OP'fLY P L 86-36 EO 1 4 c DOCID 4033692 P L 86-36 EO 1 4 c iS 6eO OPPORTUNITY U Just where does the real collection manager fit into this process Right smack in the middle with both feet U sing his experience he must teach cajole reprimand negotiate and concede all in the course of a day He is totally dependent upon analysts and collectors alike and must be guided by his NSRL His responsibility is endless his authority is elusive The collection manager is the catalyst that makes it all work -- whoever Are you a traffic analyst who despairs of finding a good TA job o Are you a risk-taker who likes independence and responsibility Are you still searching for excellence he is If your answers are yes to the above come Istill has doubts about the U Ifl necessity for collection managers I suggest that the Collection Association consider instituting counseling sessions to coaxl and any other temporarily wayward collection managers back into the mainstream of NSA thinking Ours is a noble profession no matter who does I it talk to us Regardless of grade we may have just the job for you be receiving more We're searching for a feW excellent men and women If you think you are one of them IG221 on 963-3895s 3111 86-36 E8 - EO 1 4 c P L 86-36 December 1987 IIAUQbJii VlJ E target is getting a lot of attention and will _ _ _---II P53 P L I CRYPTOLOG page 32 SE6RI r p COl n rT cgA r r b O rbY 6eo DOCID 4033692 P L 86-36 U For further information and applications contact 963-1103 r16 I ATTENTION ROCKET WATCHERS BULLETIN BOARD FOUO The History Division in cooperation with WI is undertaking a historical study of FOR CRAY USERS DEFSMAC and its predecessor organizations SMAC and the A4 Watch ff you have records FOUO Cray Research Inc has a quarterly coyeripg those orgnizations or have personal publication called CRAY Channels for users of its knowledge of significant organizationar or computer systems If you are interested in SUbstantive events please caUl lor obtaining a copy contact Henry Schorreck at 972-2355 You may also CRYSCOM exec A53 P13 963-4196 write them at T542 SAB 2 Door 3 I I ATI'ENTION LINGUISTS WITH IBM-PCs OR COMPATIBLES FOR SPA N I I-IqNGUISTSp L 8 6 - 3 6 I JPYdUO Still a favorite and available is the 19B1 U lisariol'l-lil1 clicti rY-building reprint of SPANISH VOCABULARY A and look-up tool designed for the IBM pc 0l Translator's Aid PI Language Publication$ No l compatible running DOS 2 0 or above with two disc S-220 300 1979 compiled byl It is Copies can be obtained fro -IP -16 - drives floppy fixed and or cartridge memory-resident and requires 140K of RAM in HQ BA187 Mail orders only are accepted addition to whatever your word processor needs According to the specifications it works with Word CISI Perfect MultiMate VoksWriter Delux Word Wand Palantir WorldWriter and other word SPRING CONFERENCE processors It also is said to work with pfs Write and WordStar but the latter two do not accept Theme INFORMATION SECURITY Anyone interested in accented characters 18-22 MAY 1987 experimenting with this as yet untested software I it came as a free sample in the mail is invited to GENERAL SESSIONS IpI6 963-1103S call or writ Friedman Auditorium Daily 0900-1130 and 1300-1500 t INSTITUTE ON MACHINE TRANSLATION U Georgetown University Division of Interpretation and Translation will offer an Institute on Machine Translation from August 1014 1987 Persons who had participated in the 1985 Institute may register just for the sessions on new material which will be given on 13 and 14 August Fees are $200 for the five-day Institute and $120 for the two-day sessions CONCURRENT TECHNICAL SESSIONS FANX II National Cryptologic Course Center Tuesday Wednesday Thursday Complete program with topics speakers and schedules will be published soon For further information contact one of the Conference Co-Chairmen U Topics to be presented include basic issues of MT recent developments in the field and future trends P L l 1 - 86-36 December 1987 CRYPTOLOG page 33 I dlt dl I IetAL U 6f fL' _ 2 987-7446 or P34 963-6119 DOCID 4033692 IrHE STORYOF APRlVATEGERMAN diPHERO WW I o L u J ' - - ' -' - - - - - --_- I Aug-Sep 1986 I The postcards were found in a flea market in Stuttgart in 1979 or 1980 by Tom Brousseau while we were both assigned to Europe Tom showed me the postcards and though I found the cipher not difficult the German script and language presented a problem so we set them aside In 1983 Tom lent me the cards for another attempt and while I was at it - - - - - - - - - - - -- a co-worker in A2 and a good 204 P L 86-36 o Verdun which began with a German offensive on 21 February 1916 with heavy casualties on both sides Frank had such a good time in Stammbach that he spend the next five days there not speaking a word of English P L 86-36 The niece Anna who now lives at the address of her Aunt Anna on the postcard was perplexed German linguist strolled by and soon read them about the postcards How did they end up in a flea The only problem was interpreting a phrase in market in Stuttgart she wondered As do we Martin's card to Anna which we believe reads Today I finished preparing the flower vases But there's something else that puzzles us more How was it possible that correspondence in cipher Then when Frank was nominated to study at between a soldier and a civilian was permitted in USARI in Garmisch we thought itwould be fun if time of war Granted it was a simple cipher he would go Stammbach which we learned is transparent to a professional But what about the abou t 20 miles north of Beyreuth and check the enigmatic message The German linguists who local cemetery and the records at the town hall looked over the decrypted text agree that the that But we really had no expectation of finding any cryptic passage is best translated Today I trace of Anna or Martin as 70 years had passed finished preparing the flower vases But it seems since the postcards were mailed an unlikely undertaking for a soldier albeit a horn When Frank got to Stammbach in 1986 he learned player especially in January Could it be a private that Martin was born in 1884 and that Anna was message with the meaning I'm going to the born in 1884 and died in 1949 They had no front Perhaps some reader has the answer children Among the difficulties we faced in translating the How Frank found the information and what text-the cipher posed no problem-is that the happened as a result is a story in itself To postcards were written in pencil and that the summarize briefly he simply walked along the writing in hard-to-read old-fashioned German street where his pension was located saw a grocery script is smudged Furthermore the spelling is store owned by one Karl Wirth and told a non-standard it is thought to reflect a local dialect salesclerk that he was seeking information about So other readings for the puzzling phrase cannot be Martin and Anna Wirth One thing led to another ruled out and he met a host of Martin and Anna's relatives THE CIPHER among them Martin's niece Anna who told him that Martin had died in France in 1915 The 1 2 3 4 5 6 789 postcard shows him alive in January 1916 It is A E IOU L MN R likely that Martin was killed during the battle of December 1987 o CRYPTOLOG o page 34 oa OFI 'ICIl b Yel O lIJJPY DOCID 4033692 CARD FROM ANNA I was please to receive your dear little card We are having very beautiful weather LI EBER MARTI N BIN GESTERN GANZ GUT NACH HAUSEGE KOMMEN VON NEUEMARKT BIS HElM Today I finished preparing the flower vases Best regards from your loving husband Martin 2 E KLASSE GEFAHREN BIN HEUTE GANZ AUFGEREGT r----------------------------------------------l Solution to I I I I LASSE RECHT BALD ETWAS VON DIR HOREN ODER KOMME SELBST NOCH EINMAL WHEEL OF FORTUNE I I Oct-Nov 1986 I I I I I I I I HElM MIT HERZLICHEN GRUSSEN DEINE TREUE ANNA I I I I I I I HABE DEINE UHR SELBST MIT ZUM UHRMACHER GEBRACHT I I I I Dear Martin I arrived home well yesterday I travelled second class from Neuemarkt Today I am very excited Let me I I I hear from you soon or I I I come home again With I I I I heartfelt greetings I I I I Your true Anna I brought your watch I I I I IL to the watchmaker myself 1 WILLIAM FRIEDMAN 2 ARLINGTON HALL 3 CIPHER CLERK 4 DOUBLE TRANSPOSITION 5 BOOKBREAKERS FORUM 6 BURN BAG 7 CALLSIGN ROTA 8 ROTA SPAIN 9 DEPTH READING 10 CHELTENHAM 11 ONE-TIME PAD 12 BLETCHLEY PARK 13 SPREAD SPECTRUM 14 AIRPORT SQUARE 15 PLAYFAIR SQUARE 16 NORMAL DISTRIBUTION 17 FOR OFFICIAL USE ONLY 18 COMMUNICATIONS SECURITY 19 ZIMMERMAN TELEGRAM 20 HAGELIN MACHINE JI CARD FROM MARTIN LIEBE ANNA DEIN WERTES KARTCHEN MIT FREUDE ERHALTEN BEl UNS 1ST DAS WETTER SEHR SCHON DIE BLU MENVASEN HABE ICH HEUTE FERTIG GEMACHT ES GRUSST DICH BESTENS DEIN LIEBER MANN MARTIN Because of Technical Difficulties We are Experiencing Delays in Prooucing CRYPTOLOG CRYPTOLOG page 35 FOR OFFI6IAL USE ONLY December J987 DOCID 4033692 P L NSA-CROSTIC No 63 86-36 Of all the puzzles that appear in CRYPTOLOG the NSA-Crostic is by far the favorite We're happy that in this issue we can satisfy the requests of many readers This particular NSA-Crostic has an added wrinkle Is it a clue ora bonus That depends on when you hit upon it It's a clue if you find it early enough or a bonus if it becomes evident after the fact A Insouciant 37 B Normal personnel loss 22 115 97 195 26 75 --------112 61 48 66 85 77 21 106 181 C Hybrid bramble 148 134 14 141 76 62 185 120 74 153 36 13 12 173 60 130 168 127 149 189 72 100 160 47 183 131 11 108 103 30 92 128 64 50 52 84 56 '93 160 118 138 143 63 104 67 40 86 133 159 179 38 57 73 182 162 109 90 '61 20 176 88 16 158 147 71 105 42 124 166 137 70 79 89 163 69 43 99 121 94 '9' 177 33 17 82 49 25 126 184 68 65 28 80 170 18 190 93 114 169 7 181 111 44 194 102 119 186 15 51 91 172 146 58 125 150 35 136 154 151 129 167 31 155 19 98 10 135 34 53 174 175 18 107 156 192 123 54 188 46 24 178 122 83 95 165 32 142 101 145 157 113 96 140 152 116 59 87 45 139 81 55 23 164 27 132 11 144 D Mu's neighbor E Reddish bird dog 2 words F A family ruler G Diligent H Innate I Syncopated 29 J Fortify K Essay l Supply fully M Patron N Crowning achievement O Casual P Plane figure not a Greek conflict Q On edge R Apartment S - - del Vaticano T Seminary subject U Victim of Paris V Water nymphs not dryads W A cryptanalytic phenomenon such as IT medieval garb X Amateurish December 1987 CRYPTOLOG page 36 FOR OFFIGIl L USK 01'lLY 41 39 110 DOCID 4033692 1H 2 C 3A 4 E 5 J 6 L 7 Q 8V 9 T 10 R '3 0 14 C '5 P 16 G 17L 18 N 19 R 20 F 2' 8 22 A 26 A 27 X 28 N 29 1 0 H 1 R 32 U HL 345 35 0 360 38 C 9 R 40L 41 X 42 H 4 J 440 45 W 46 T 47 G 488 50 1 52 1 53 54 T 55 X 56 J 57 C 60 E 61 8 63 K 641 65 N 668 67 L 70 1 71H 72F 75 A 76 C 778 81 W 82 L 83 91 P 92 H 1020 103 H l04l 111 X 12 B 113 V 114 N 121 K 122 U 12 T 124 H 125Q ' 4 C 135R 136 Q 137 I 138 K 144 X 145 V 146 P 147 G 148 C 149 F 150 Q 155 R 156 T 157 V '58 G 159 N 160 J 11 1 F 166 H '67 R 176F 177 L 186 p 187 8 59 w 133 M u 84J 93 N 132 X Puzzle fans are invited to try their hand at compiling an NSA-Crostic with or without additional features For helpful hints and guidelines consult the Puzzle Editor December 1987 o CRYPTOLOG page 37 FOR OFFI6IltL USE OPt 'LJPY This document is from the holdings of The National Security Archive Suite 701 Gelman Library The George Washington University 2130 H Street NW Washington D C 20037 Phone 202 994-7000 Fax 202 994-7005 nsarchiv@gwu edu