mGBPDUV0mGBPD11 w v lBl W lBl $ BWWVV D1 sUl B lB'I WI UlIDI fJ WUl l1GBPDIl1ID lBrn IlUlIJl1lIJrn OOGJ 1JEJf2 P L '1 86-36 h CONFESSIONS OF A BRIEFER U oooooooooooooooooo d o o o o 1 STILL MORE ABOUT PASSWORDS U ooooooooo 5 CRYPTOGRAPHY AT GLOBECOM 82 U 7 11 REVIEW DIGITAL TELEPHONY U CUMULATIVE INDEX PART THREE KEYWORDS U oooooooooooooooooooooooooo oooooo 13 OUT OF MY DEPTH U oooooooooooooooooooooooooooooooooooooooooooooooooooooooo 33 THE INTELLIGENCE WATCH OFFICER U ooooooooooooo 35 COMPUTERIZING TRAFFIC ANALYSIS U ooooooooooooo l 36 NSA-CROSTIC NO 47 U oooooooooooooooooooooooooo David H Williams oooooooooo 44 I I 1 'fIllS BOCUMENT CONTl INS COQEWOKQ t UTEKIl J L o o QW SSIl IHB 8Y JrSl' esslf 1ita a QkF SSIPJPY QN eri ifta ift A eRSJPY's ge eEMiRa i8ft Declassified and Approved for Release by NSA on 10-12-2012 pursuant to E O 13526 MDR Case # 547713 Re iree aCID 4009861 Published by Pl Techniques and Standards Editoria' VOL X No 5 MAY 1983 PUBLISHER BOARD OF EDITORS Editor Asst Editor Production 968-8322sY ---------l 963-3369S 963-1103s Co11ection oooooooooooo 1 Computer Security At my bank having direct mail deposit qualifies me for certain privileges but when I applied for them I was told I was NOT a direct mail depositor I disagreed1 my paycheck is sent directly to them and has been for some time Then I discovered the key my account is joint with my wife and the computer lists her name FIRST So when they retrieve data about MY accounts oooo 963-3961sy ------------ 1 968-73 13S Cryptolinguistics Data Systems Informa t ion Sc ien c e 963-1103s 963-4953sY --- 1 963-5 711s Mathematics ooooo J I 968-8518s Puzzles oooooooooo David H Williams H963-1103s Special Research ooooo Vera R Filby 968 7119s 1 For subscriptions send name and organization to CRYPTOLOG Pl or call1 1963-3369S P L 86-36 To submit articles or letters via PLATFORM mail send to crypto1g at barlc 5 bar-one-c-zero-five note no '0' in 'log' Contents of Cryptolog should not be reproduced or further disseminated outside the National Secur ity Agency without the permission of the Publisher Inquiries regarding reproduction and dissemination should be directed to the Editor Some years ago when I dabbled in real estate I had several fat listing books and kept them in order by area price and number of bedrooms The listings were typed and often contained errors Some were easy to spot If the tax district or the number of bedrooms was wrong it was easy to make a pencil change on my copy and move the listing to the right place in the books I notice that real estate agents are computer ized these days They go to a terminal key in a price range number of bedrooms geographic area etc and out comes a string of listings Who makes the corrections I doubt that anyone does It wouldn't surprise me to hear that some real estate people hide new listings when submitting them to Multiple List data bases by putting typos into key data fields to give their own agents the first shot at a new listing Between accidental and intentional typos I wonder how much of the existing data actually gets to the requestor Please don't write in and tell me about good programming practices That's not my point Both we and our targets are coming to rely more and more upon data retrieval for our information Anybody who can read and type can call up data on most of these systems but only the more innovative people will be able to squeeze out of the system the hidden data Knowing how to do this could depend upon understanding how people behave when they use a data base and also how the data base itself really works It could in fact become a new cryptologic skill fi PQA QPPI8IhEJ eSB 8lft r aCID 4009861 peR epP e AfJ eSB SUM' P L l ake up Jim 86-36 The briefing's over rA I Huh what did you say Oh I guess must have drifted off Sure is A hard to stay awake in a dark br iefing room after lunch Say Fred did you get copies of that guy's slides for me to read back in the office so I can find out what Project RATTLECAN is all about Does this exchange sound familiar If not then this article is not really for you Keep on reading though you might learn something that will be useful later But if you are one of those who have been bored to sleep on numerous occasions as some well-meaning project officer or analyst read an endless succession of slides to you perhaps this article will contain some thoughts that may help you avoid inflicting similar boredom on others It may even make your briefings more effective One of the most basic causes of poor briefings here at NSA and elsewhere as well I am sure is the mistaken view that briefings are a good concise way to transfer information to people Absolutely nothing could be fUrther from the truth Yet we constantly use briefings to br ing people up to speed on a wide range of detailed and complex topics The subtle deceit of briefings is particularly interesting when someone has received a Good briefl from his audience Just ask the br iefeeafterwards how many tanks the Zend ian Army has or how many communications circuits are at Field Station Xapa or some other detailed question on the topic and most likely you will get some answer like Well I don't recall exactly but I'll call Tommy Talker who briefed me on it and he'll have the answer In this case Mr Br iefee may not have the facts but he did get the message The fallacy of using briefings to bring people up to speed is that briefings are not a good medium for the presentation of a lot of objective factual data They are however an excellent medium for affecting people's atti tudes and emotions In our society we have so thoroughly suppressed our emotions e g Grown men don't cry that we would tend to deny that we are even subject to a play on our emotions especially in some intellectual palace like the National Security Agency UnfOrtunately this leaves us extremely vulnerable to approaches from a nonintellectual angle and the briefer who discovers this vulnerability either by accident or as an active intellectual discovery can use this approach to produce consistently good briefs This is a fact long known to Madison Avenue and it is equally true here at the Agency When you have had a good briefing what are the things that you most readily recall In all likelihood it will include items such as the command and presence of the br iefer and the quality of the graphics You will remember the organization that sponsored the briefing and those memories will be of a competent and professional outfit You will know that the topic briefed is very grave requires immediate action by your organization C is in competent hands or needs more resources in the out years But you will not recall specifically Why you have those feelings May 83 CRYPTOLOG Page 1 ern 4009861 FQR QFFISlhlO elSB eUlOY All this having been said then how can we use this knowledge in building a good brief The most basic step is to decide what emotional message feeling or attitude you want to inculcate in the listener Generally the attitude or emotion will be one favorable to the topic and organization presenting the br ief For instance a project officer will generally want to leave listeners impressed wi th the importance of the project so that future requests for support will be favorably endorsed In short although briefings are generally thought of as being informational most briefings are sales pitches Once the basic message of the briefing is established all other work should support that message In briefing there is no substitute for the competent briefer a person who speaks with all the self-assurance of a Nobel Laureate but who at the same time compliments and involves the listeners one who does not preach to them For some people these qualities are natural but for most they can be acquired The secret is lots of practice and the knowledge that you know more about your topics than your listener Armed with this confidence the briefer should never use a script If you know what you are talking about you don't need a script If you don't know what you're talking about you shouldn't be briefing Bishop Fulton Sheen never used any notes on his very successful TV show during the 1950s When asked about this once he cited a remark he had heard as a child An old woman walking out of church was complaining because the bishop had read his sermon from notes Her question which impressed Sheen - was How the devil can he expect us to remember what he's saying when he can't remember it himsel'l FIGURE 1 I CORPS III IV V ABN CORPS CORPS CORPS CORPS TOTAL Freedom from the script allows the briefer to give the appearance of being extemporaneous as he or she responds to the quips comments and questions of the briefee It also permi ts the br iefer to tailor the br ief readily to the background and interest of the listener Finally the freedom from the script allows the briefer to establish a pace or rhythm for the briefing that enhances the authority and assurarice of the presentation with the script eliminated the only tangible form left to the briefing is the graphics These are as critical as the presence of the briefer and in fact can often make the briefer seem more professional and the briefing come across better In the choice of graphics it is especially important to keep in mind the emotional appeal of the brief The Chinese writer who said that one picture was worth a thousand words knew what he was talking about and the proper choice of graphics can save a lot of talk and questions For example consider the graphics in Figures 1 and 2 Zendian Army Power Figure 1 is just a compilation of numbers The briefee will read it rather than listen to the briefer--and you'd better hope that the br iefee doesn I t have some spur iou knowledge or he might make some comment like I thought the Zendians had Type Q tanks instead of Type Ys This sort of question could well be enough to throw an inexperienced briefer off pace or worse yet lead to intellectual questioning of every statement Figure 1 has has yet another critical flaw In column 1 the numbers don't add up to the total shown If the listener notices this he she will spend the rest of the briefing adding up any numbers that appear looking for other errors ZENDIAN ARMY POWER TROOPS 55 000 TANKS 250 HEAVY ARTY 200 MED ARTY 500 APe 600 HELOS 45 58 000 61 000 58 000 25 000 300 312 270 212 220 220 550 560 550 200 600 500 550 43 48 45 350 247 000 1132 632 2360 2250 531 There are no units in the Zendian Armed Forces with the designator 2 or II because the Zendians consider this number bad lUCk There are no tanks per se in the Airborne Corps but there are approximately 200 of theso-called y-type tanks that are iil reality a lightweight high-speed tracked anti-tank gun May 83 CRYPTOLOG Page 2 Figure 2 wager - I - l - Figure 3 May 83 Page 3 aCID 4009861 Figures 2 and 3 previous page are good examples of effective graphics Both of them evoke a strong emotional response In fact either of these pictures could well move the listener to a rendition of war stories about when he she drove tanks rode in helicopters or had some related experience This will imbue feelings of camaraderie between the briefer and briefee that will make the briefee much more amenable to the briefer's message These two pictures have some other practical advantages they will serve to cue the briefer they will not limit the briefer may wish to make remarks they don't contain the intellectual snags that are found in Figure 1 and perhaps best of all they will not have to be changed or updated unless the Zendians get rid of that tank or helicopter By concentrating on influencing emotions or feelings we do not ignore the facts We just use them in a different manner Rather than being sort of inert things the facts that we have to use are woven into the fabric of the brief in a way that supports the basic message One way to do this is with amazing facts We all have a store of amazing facts but probably don't realize it For example the fact that the Zendian Navy has 89 operational submarines will probably mean very little to anyone but an avid naval buff but the same fact cast in a different context becomes an amazing fact The Zendian Navy has the largest submarine force in the third world Bar pie and line charts are all effective ways of presenting amazing facts such as this While you never want to read your slides to your listeners there are times when you may want to let your listeners read the slides themselves This provides a change of pace for both br iefer and br iefee It's a quick way to slip over what otherwise may be a long narrative and it involves the briefee in the briefing process more actively Such graphics should never be long textual passages The proper form is short bullets ideally only one or two words each See Figure 4 Figure 4 The briefer can introduce this type of graphic wi th some line such as These are the characteristics of ooo whatever the subject is The briefer should watch the faces of his audience and move on to the next graphic as soon as the the expressions of the listeners show that they have read the graphic the When using slides like this however make certain that it is really a picture of what you say it is or it will be almost as bad as having numbers that don't add up May 83 UNITARY PLANNING I W EMPHASIS SURVIVABLE COMMUNICATIONS COMPUTER INTEROPERABILITY PRODUCTION ENHANCEMENT COMBINED WORKFORCE In summary the key to effective briefing is to remember that br iefings should be used to form attitudes or affect emotions not only to transfer objective facts The effective brief should have one central underlying attitudinal or emotional message that it is attempting to deliver and all aspects of the briefing must support this The briefer should not use a script and the graphics should be simple and chosen with an eye to their emotional impact Facts used in the briefing are much more effective when placed in some sort of comparative context Slides to be read should be read by the br iefee not the briefer Good luck Good brief Epilogue In closing I believe it is necessary to comment on the ethics of briefing It is evident that armed with information about the weakness of our psyche an unpr incipled bureaucrat can take considerable advantage of his or her colleagues At present the only sure counter to this is the individual integrity of the briefing organization which must use its power only for pure motives Unfortunately a full discussion of the ethics of briefing is beyond the scope of this article but perhaps some reader may feel an urge to expand on that topic CRYPTOLOG Page 4 peR epPIQIM JSII QIJI Y PSR SPPieUt'EJ 8BB Stu 4009861 4f MORE ABOUT PASSWORDS U _ _ _ _lpI3 1 o L 86 36 A USER VIEWPOINT The p ssword controversy continues On the One hand I I and others have stressed the need for greater security to avoid potential compromise On the other hand 1 nd others have made a strong case for short easy-to-type and memorable passwords to avoid needless errors and frustration on the part of the user I am in sympathy with both causes although I violently object tol Iclosing paragraph CRYPTOLOG Mar 83 p 38 he states that an easily remembered password is easy to type regardless of length and that he doubts that a non-typist i e one who does NOT find a long memorable password easy to type will use a terminal for very long I would argue just the opposite The probability of error rises progressively with each additional character added to the password especially since characters are not being echoed on the screen Also my personal observation has been that the majority of non-secretarial users-linguists programmers and managers among them--are in fact either non-typists or poorto-fair typists myself included yet terminal usage among these groups is increasing rather than decreasing as more and more people become aware of the advantages which a computer terminal can provide to the professional user Having said this let me state that I think the problem is easily solvable in a manner which should satisfy both viewpoints Instead of viewing the problem theoretically--short passwords are good but breakable large alphabets pass phrases and passcodes are good but will result in higher error rates-we should take the Agency environment into account Many Agency computer systems have a feature which automatically kicks out anyone who unsuccessfully tr ies a user ID-password combination three or four times in a row Those which do not can be easily modified to allow this capability If the office security manager is alerted to terminals on which three unsuccessful IO-password combinations have been tried I doubt very much that a hostile entity would have any success in breaking even a five- or six-letter single-case mnemonic password before being apprehended At the same time a poor typist gets several chances at entering the combination correctly before setting off the alarm This system thus provides the best of both worlds a user-friendly password environment which is for all intents and purposes immune to exploitation P S A challenge to all UNIX users how many of you can type Low flying bees eat wax beans to usel lexample of a pass phrase and get it right thefirst time It's easy to find out set termirlal type to STTY -ECHO type in the phrase enclosed in single quotes Le 'Low flying bees eat wax beans' and hit RETURN A system message will appear as follows flying bees eat wax beans not found or whatever you REALLY typed After you're done experimenting STTY ECHO will make your characters visible again At an informal testing in the KEPLER laboratory I got the phrase right 3 out of 7 times however I fall under the non-typist category THREE GOOFS AND YOU'RE OUT On the system I use a user signs on with user initials and then is required to type tn a password If the person fails they may try again and again and again To prevent exhaustive searches why not flag to the sys terns operator any console that tries say thre times Or lock that terminal out May 83 CRYPTOLOG P L Page 5 aCID 4009861 PASSWORDS FRIEND OR FOE letters of the last name A central registry is being established to resolve conflicts Speaking of passwords what is an acceptable balance between security and convenience I have found what I believe to be a rather comfortable and simple solution but then I'm sure SOMEONE will disagree I simply choose a word o o such as walnut then alter or mispell it wailinut wawlnut o o etc Not enough to make it difficult to remember but definitely making it more secure and harder to guess MORE ON PASSWORDS AND HUMAN FACTORS from COMPSECNEWS June 1983 One of the frequently heard complaints from persons who changed to more clever passwords containing special symbols or mixtures of upper and lower case was that when bringing some of the terminals up cold and prior to loading the terminal emulator they could not login True that each keycap may not send the same character before an emulator load as it will after but when those infrequent occasions arise you can still prpbably login if you only knew what keys to press For example one system I know of uses the back tab key to produce the I prior to emulator load Obviously an inconvenience but most users do not regularly have to perform initial terminal loads Other systems that I have seen have accounts called LOAD These accounts do not require the person who logs in to enter a password only to choose a desired emulator After the selected emulator is loaded the real user login is then required If your system does not have this feature complain to the system Guru Incidentally notice that an emulator load is all that the LOAD account can do Another comment had to do with the login-id being secret The suggestion centered around doing away with initials as the login-id and using a secret account name in addition to a secret password Maybe a useful idea but one point missed by this comment was that the current login algorithm used on most of the systems does not reveal what is wrong when the login fails For example notice of login failure only appears after the id and the password has already been given Another problem with this suggestion is we already have a requirement to identify individual use of sensitive computers Work is currently underway to assign each and every user of our computer systems a unique id that will be the same regardless of which computer system is used The adopted format will follow first two initials followed by the first five L - ----JI June 1983 Keep your comments coming In particular I would like to see more comments from users about the consequences for them of various access restrictions password procedures etc How have some practices on the systems YOU use hindered or helped YOU in your work I know that it i sfunforal t fxouingeniouspe pIe out there to think up new password schemes and gimmicks but the computer security experts are pretty inventive and ingenious too What they need more than new techniques and ideas is some clear feedback from users about the COSTS and BENEFITS of different kinds of procedures currently in use If they get a clear indication from users that certain methods of implementing access restrictions impose a relatively high cost on users they will be motivated to use their ingenuity to find other and better methods that are just as secure but less costly to the user I was interested to note in the COMPSECNEWSitem above the assumptions that 1 users rarely had to load terminals cold and 2 coping with keys that send different characters before and after emulator loading constituted a minor inconvenience at most What do you think about those assumptions As a hunt-and-peck typist I find odd-ball special characters are a MAJOR stumbling block even when the keycaps agree wi th the character that gets sent Even the shift key and CTRL key are frequent error-makers for me Also my impression is that users have to down-line-load Delta Data terminals more frequently at least for some systems than the COMPSECNEWS editor assumes In fact if I had to cope with that problem even having to load a terminal once a week would be far too often for me The LOAD account mentioned in the article seems to me a much more promising and user-fr iendly approach than if you only knew which key to press For a programmer or someone thoroughly familiar with the terminal and software perhaps keys that send different odd-ball sets of characters at different times may present only a minor annoyance in fact computer folk seem to thrive on and positively welcome such problems Most others however are very unlikely to agree with them Computer specialists and computer security specialists in particular need to be reminded that situations constituting brief nuisances or even amusing challenges for them can be stressful and exasperating for other kinds of users May 83 CRYPTOLOG Page 6 FeR eFFl'el'A ' 8SB eUI ll -M E D P L P L 86-3 aCID 4009861 Cryptography at GLOBECOM 82 P L 86-36 Review Digital Telephony U DES-type algorithm using 800-bit vectors in place of the 48-bit vectors of DES was proposed at GLOBE COM 82 the 1982 Global Communica U tions Conference for the encipherment of medical records The key would be 160 900-bits long The Belgian author of the paper Desmedt claimed that this would protect medical records during the life of the person concerned In reply to a question he admitted that he did not know how to keep the 160 900-bit key itself secure and intact for the 100-year period U The Desmedt paper on super-DES was one of five papers on cryptography presented at GLOBECGf 82 in December In addition there were five other sessions on coding primarily speech coding which proposed reducing bit rates for video and voice and facsimile Compression techniques and the ability to recover from channel errors are critical to the use of digital encryption techniques U One of the surprising papers was about a detailed experiment with analog encryption at Bell Labs Apparently low bit rate encryption causes so much loss of voice quality especially over low-quality lines which cannot support 9600-bps rates that there is a growing demand for encryption which sends analog waveforms The Bell Labs work has been done by computer processing but they expect to develop a real-time circuit after which their VLS I chip designers will examine the cost of single-chip analog encryption That could have a revolutionary effect on secure voice and on cryptanalytic priorities May 83 U Two other surprises were the sophisticated insight into the strengths and weaknesses of various public key schemes especially the flaws of the Hellmann-Merkle algorithm and the importance of the recent Racal-Milgo patent on finding large primes for the RSA public key algorithm Several speakers stated that the RACAL-MILGO algorithm had made the integration of DES and the RSA algorithm feasible as the basis of a switched ad hoc public cryptographic network U This demonstrated interest in the feasibility of the RSA algorithm as a means of keying DES links is more interesting in the light of the Inman interview Science Dec 82 which identified RSA as a secure method U The leadoff paper by J Michael Nye a self-styled cryptographic expert described the methods and cost of intercepting telecommunications in the US and gave a list of 26 domestic cryptographic suppliers offering 104 products and 13 foreign vendors offering 81 products in the US market The list of suppliers and products is growing and the impact of Personal Computer encryption is yet to be felt This is a very big change from ten years ago when only a few companies supplied cipher equipment to the US market Most of the products are for fixed telephone service but as the new technology of cellular radio develops over the next decade the market for voice encryption to protect the 900-MHz mobile circuits from interception will increase to millions of vehicle radios CRYPTOLOG TOP l iQgBT Page 7 HUBR-A aCID 4009861 U Cox Jayant and McDermott of Bell Labs gave a paper on a time-frequency segment permutation analog encryption which they believe is very secure against cryptanalytic attack without loss of voice quality or syllable intelligibility The delay for the scrambling and descrambling is no more than 256 to 512 milliseconds for 16-msec speech segments Each 16-msec segment sampled at 8000 Hz to give 128 samples is converted to sub-bands by digital filtering and the sub-band vectors are then permuted The digital vectors are stored in a buffer with a memory capacity up to 512 msec A cryptographic keystream decides on the segment permutation and also decides which time segment will be sent Any time segment can be delayed up to 256 msec The input test data were voiced digits in random order used to avoid the redundancies of normal conversational speech U The cryptographic scheme is to fill the buffer with 16-msec segments then send all of them in some pseudorandom order until the buffer is transferred to the receiving end Then the buffer is refilled and the scrambling and transmission begin on the next multisegment block of speech It is not clear from the published paper whether the segment transposition key is the same or changes from block to block are used to establish and maintain sample timing and are also used for channel equalization since they give the impulse response of the channel U The main thrust of the designers' work They now was to get good intelligibility want to reduce the software algorithm to a hardware device and VLSI design can follow from there The audience showed substantial interest in this scheme U Despite the rapid introduction of digital channels for many years to come most of the world's telephone connections will be made over copper circuits that will not support high bit rate digital speech One of the driving forces behind the development of digital transmission and local loops is the desire for high-quality secure speech If the hybrid analog-digital scrambling gives good enough speech quality and no particular security weaknesses become known the market pressure to develop digital services to the 64 OOO-bps level ION Integrated Digital Network may reduce since customers won't have to buy them and this could affect an important part of the digitization of the telephone network U The scrambler was implemented on the BTL Digital Signal Processor and tests showed that it gave better intelligibility than simple frequency inversion scrambling The scrambled signal envelope sounds like birds chirping To maintain synchronization a series of pulses is sent down the channel whenever the scrambling buffer is reinitialized and these high pitched pulses send like cricket chirps interleaved with the signal U Although the system is described as an analog scrambler it is clearly a 5-stage analog-digital process in which most of the processing at each end is digital but the transmitted signal is an analog waveform It has been tested over the Murray Hill phone system with addition of simulated white noise and phase roll a channel impairment U Because sample-to-sample fidelity is important to speech reconstruction i e the sampled speech at the receiver must match the samples that are supplied to the digital-toanalog D A converter at the transmitter it is necessary to equalize the channel to compensate for channel distortions and to synchronize the instants of sampling at each end The synchronization pulses cricket chirps May 83 CRYPTOLOG Page 8 iii EO 1 4 c P L 86-36 aCID 4009861 'l 8l S6SR6'l ijUBRlt This is undoub tedly a reflect lgn lcrf'l tHe lr apid transborder flow of techn e 1 i 1'i1oMtion between academicians and the arrival of increasingly capable people into the arena of public cryptology Without doubt the technical quality of the work will increase and will threaten SIGINT U The particular cryptographic scheme that Desmedt et al propose is a version of DES in which blocks of data of 1 600 bits are enciphered under the control of a 160 900-bit key The S-boxes of DES are replaced by oneway knapsack functions This revised tIS box instead of operating on eight bits operates on 200 bits and there are eight of them operating in parallel Each of the new s boxes is initialized with 100 integers of 20 bits so that it contains 20 000 bits There are eight s Boxes which use up 160 000 bits of the key Because of the trapdoor function even if the 160 000 bits were known and all the S-box outputs were known it would still be very difficult to compute the BOO-bit input However the 160 ODD-bit key is not known It is kept secret That makes it even harder to compute the input from the output The 200-bit outputs are expanded up to 20B bits and then hashed down to 100 bits to give the BOO-bit output This complicated process is iterated a number of times A stream or block mode with this algorithm is possible U The Desmed t paper on Super-DES began with a critique of existing cryptographic methods interleaved with some familiar commentary on NSA intervention in the DES design A point of interest is the statement that the Geneva Management Group in 1981 concluded that DES was not adequately secure Desmedt argues that encryption algorithms that iterate the basic operations many times provide higher security than the individual operations e g substitution transposition but are impractical to implement on VLSI chips He also argues that a DES-breaking special machine may be costly today but in 20 years could be cheap enough to break messages enciphered on DES now He also acknowledges that no shortcut solution to DES is known Hellmann's insinuation that a trapdoor was built into DES by NSA is referred to The problem of public key algorithms such as the Merkle-Hellmann scheme is described as either they have known weaknesses or they may have unknown weaknesses What is notable about all of this critique is that the authors are professors of mathematics in Belgium and they are very up-to-date in the state of cryptology in the public domain One of them spent 197B-79 at UCLA Berkeley doing pertinent research May 83 CRYPTOLOG Page 9 'l BP SB6ftE'f tHf8ft'A EO 1 4 c P L 86-36 EO 1 4 c DOCID 4009861 'f1't6 lf S I e R I 'fMU'ftJtlfjfft tl Rtlfrt L o 8 6- 3 6 Lrresistible from both intellectual and marketing viewpoints D Yiu and Peterson of Hewlett-Packard gave a paper on a single-chip VLSI Public Key algorithm The algorithm is Hellmann's discrete exponential scheme using Galois Field arithmetic The chip has 12 000 transistors and is designed for a 441Hz clock rate The Public Key algorithm would be used in conjunction with DES to distribute keys for the DES algorithm The purpose of the chip design was to give higher speed and lower cost for encryption The use of Galois Field arithmetic eliminates the need for carry or borrow operations and the arithmetic operations can be executed by linear feedback shift registers The developers expect to use the chip in a computer network but the company according to Yiu has no commitment to market it D Doctor Yiu mentioned the recent Racal Milgo patent for finding large prime numbers in a few seconds as an important breakthrough in implementing Public Key networks The patent No 4 351 982 claims that it reduces microprocessor computation time to find a set of 200-bit primes from 1 200 hours to two hours Desmedt stated in his talk that the RACAL MIL GO datacryptor took only 17 seconds to distribute key but it was not clear that it would find RSA primes in that short time Hollander of BTL Bell Telephone Labs has a patent application that purports to find large primes very quickly The Japanese are developing a chip that will do RSA encryption at 50 000 bps The work is a joint project of NTT NEC Hitachi Fujitsu and Oki Sandia has developed an algorithm that will do multiplication modulo C in log2 C 7 clock pulses which is an improvement over the conventional modular multiplication which takes N x N clock pulses for an N-bit modulus The Sandia method would take only N 7 clock pulses It is aimed at RSA encryption using 512-bit prime numbers At 2041Hz clock speed they expect to be able to encrypt at 25 Kbp s Now that the RSA Public Key algorithm has been publicly identified by a former NSA Director as secure there will undoubtedly be intensified work to make it easier to use to set up DES links The high utility of being able to dial up any other party and set up a secure link without prior key distribution is May 83 TQP D A paper on the AMI DES chip was given by Brown an AMD executive This was followed by some discussion of the merits of DES Both Yiu and Brown felt that DES was in fact secure and that the criticism that Desmedt and other authors had raised were not supported by any facts No one had been able to read or exploit DES and it had the advantage of a standardized tested algorithm The AMD DES chip was capable of 1 7 Mbytes sec so that it could be used for disc controllers It cost $75 for a single unit D Another paper at the cryptographic meeting was an Italian scheme by CSELT for a robust 4800-bps speech coder Audio tapes showed it to be resonant and of low quality The paper did not seem to offer any important new work However the topic of speech coding and other compression coding was treated at five other sessions There were 31 papers on voice and image compression of which 16 were by foreign authors The sessions were J A8 Low bit rate speech coding B6 Image processing J D4 European videoconferencing E6 Advances in speech coding F7 Speech processing D The interest in compressed speech in sessions A8 E6 and F7 was initially to allow narrowband encrypted voice signals Now a number of other applications including lowcost bandwidth conservation and interim storage of voice have emerged from the capability to compress speech D The interest in compressed video is for both teleconferences and private TV broadcasts e g Pay-TV TV relay by satellite and Direc t Broadcast Satellites DBS The common carriers and the broadcasters both see commercial advantages in being able to send TV signals tha t can be securely encrypted Because bits sec cost money the customers want the pictures compressed D In Europe there is a multinational effort to develop a standardized videoconference system with a standard video coder Some of this is for satellite applications to thwart interception but most of it will CRYPTOLOG SE Il 8T Page 10 BUBFb z aCID 4009861 'SF SI3EURRl3' SUBR probably pass overland on radio relay and optical fibers The compression allows cheaper conferences and security and the standardization will allow the Europeans to intercommunicate and keep US companies out of the equipment market U There are now hundreds of papers published on vocoders fax coders encryption video coding etc Many of the papers are foreign but the Europeans in particular have been handicapped in the speech area for example by the lack of specialized journals which consolidate the work As as result they look to the US journals especially the IEEE publications as the focus of the current work This also makes it difficult for us parties to keep up wi th the foreign work because it is spread across a number of journals and is often published in German French Italian Swedish Japanese etc However the foreign literature is growing and will become a more important source of new work in cryptography and coding U The Europeans arrlvlng on the scene after the US has identified the problems and paid for the basic research will be able to converge on coding and encryption standards to serve many of their PTT plans without the competition confusion public controversy and divided purposes that have arisen in the US in both Government and civil coding and encryption By the end of the decade they may have passed the US in these fields just as they have surpassed the US in a number of other selected technologies and industries References 7 Inman Bobby Science V218 24 Dec 1982 p 1290 8 NTIS PB-806860 Vocoders 1970 82 abstracts from the Engineering Index Apr 82 Contains foreign vocoder abstracts In P1 library 9 NTIS PB-82-806852 Apr 82 Citations from NTIS data base ture In P1 library 10 NTIS PB 82-863671 Mar 82 Facsimile Communications Citations from INSPEC data base Contains many foreign literature references Digital Telephony by John C Bellamy Wiley 1982 -tet-Is plesiochronous a familiar word It soon will be if you are concerned with international digital networks A plesiochronous network does not synchronize the network but merely uses clocks at each node that are accurate enough to keep the bit slip rate low enough not to interfere with operation The US domestic digital network is synchronized to save the cost of the node clocks but the CCITT has established clock standards to interface different national networks by plesiochronous gateway connections Because the national networks run at slightly different rates land COMSEC will have tqanticipate this plesiochronous struc e 1 4 c I P L 1 GLOBECOM 82 Conference Record 2 US Patent 4 351 982 RACAL MILGO 3 E F Brickell A Fast Modular Algorithm with Application to Two Key Cryptography Sandia National -Labs -- Re DE-AC0476DP00789 4 Simmons G and M Norris How to Cipher Faster Using Redundant Number Systems Sandia National Laboratories Aug 1980 5 Miyaguchi A S Mizuta K Furuie T Tsujiguchi and H Mizuta Combination Type Computer Circuits for Code Handling NEC 1981 6 NTIS Search Vocoders V1 2 1982 May 83 Vocoders US litera- 86-36 U Network synchronization schemes are just one small part of John Bellamy's new book on digital telephony The author who received his PhD in EE in 1971 worked as a manager at the Collins Division of Rockwell International in transmission systems then as a member of the technical staff at Arthur A Collins Inc the R D firm that hived off from Collins Radio when Rockwell took them over He is now an R D manager at the Communications System Division of United Technologies so he has substantial practical experience with modern telecommunications engineering It is notable that Collins builds digital radio equipment so well that Western Electric dropped some of its own projects and bUyS from them In addition to his industry exper ience Dr Bellamy has been an Adjunct Professor of Electrical Engineering and Computer Science at Southern Methodist University since 1976 so he is used to CRYPTOLOG Page 11 'J'8F SJ3EURftI3'f 91IBR'j 'Pep Se8RB'P SfIBFls 4009861 organizing simplifying engineering technology and teaching U His book is a well-written tutorial on telephony and digital telecommunication Although digital transmission was developed with computer and data traffic in mind the main traffic volume will be voice for a long time Analog transmission and networks will also be around for a long time but digital telephony is of particular interest where encryption is wanted as he says on page 75 because of the paucity of good analog encryption U The book covers digital networks from a general overview of the analog network in the US through voice digitization algorithms PCM DPCM APC vocoders etc fundamentals of digital transmission switching and multiplexing through dig i tal modulation network synchronization and control to high-level descriptions of several digital networks and a discussion of the future of digital telephony Among other things the book has a 16-page glossary of pertinent terms such as bur st isochronous Centrex service despotic network elastic store HDB3 code justification ratio mesochronous muldem permanent virtual circuit plesiochronous robbed digit signaling state store transhybrid loss traveling class mark and waiting time jitter These terms are indispensable refinements in educated discussions of networks This glossary will soon be in the NSA terminology base U Some points of interest in the book One of the principal difficulties in making the large telephone network operate is the variety of signals and signaling func'tions all of which have to be translated by interfaces or made compatible Common Channel signaling in the Bell System is highly centralized making the entire network vulnerable to failures in the CCIS packet traffic If a CCIS node fails to store and forward certain network information correctly the network will gradually lock up because disconnects do not occur automatically J Digital microwave radio is cheaper than T-carrier for distances as short as eight miles and the major impetus for digital radio has been the introduction of digital switches not the demand for digital traffic which can pass over the analog network May 83 Fifteen different types of digital central office switches are in service in the US and more are expected as foreign suppliers enter the US market The disadvantages of digital networks are increased bandwidth AID and DIA conversion time synchronization topologically restricted multiplexing and incompatibility with the large analog plant Voice digitization nominally 64 000 bps can be as great as 400 000 bps for specialized services such as broadcast transmission About ten different speech coding schemes are described U In the chapter on digital switching the advantage of time-domain switching is shown to be the ease of getting switches that don't suffer from blocking There is quite a lot of information in the switching chapter The chapter on network synchronization examines many of the timing problems and the solutions such as bit slips elastic stores pulse stuffing and packetization The chapter on digital networks leads from ARPANET to the ISDN integrated services digital networks Circuit-switched nets are shown to be more efficient for voice transmission than packetswitched nets but digital speech interpolation can increase the effective number of channels on a trunk if the circuit switching is fast enough Current circuit switching can operate fast enough so that even the beginnings of syllables are not lost The final chapter is on traffic analysis as the traffic engineers and designers do it U The book is well written and well illustrated with references and exercises Because of the integration of transmission and encryption as well as the continuous growth and switchover into digital networks in every country a knowledge of the engineer ing and technology of digital networks will be an essential part of a cryptologic repertoire Bellamy's book is a good beginning to this knowledge The enciphered speech cryptanalytic experts in R52 ordered desk copies after reviewing the book U Summing up Digital Telephony is timely gives good coverage of digital networks and should be a useful text and reference for some years It costs about $50 CRYPTOLOG Page 12 aCID 4009861 The following cumulative index of CRYPTOLOG Vols I through IX 1974-1982 was produced using UNIX PINSETTER tools on MYCROFT and BARDOLPH1 The index is in three parts and is being pUblished in three successive issues Part one is an index by author part two is an index by title and part three is an index by keyword Items in multiple issues February-March 1975 for example are indicated by the first month i e by Feb 75 o CUMULATIVE INDEX EQ 1 4 c P L 86-36 I -'Three Keywords 86-36 Arne rind Sep74 Language in the N S LI MAS Feb 82 Gayler - AAAS 1982 -L Feb 82 Software - AMS 19B2 1__ __ ABNER Aug N 77 Early Days in NSACQmputing Ch u f IAf1 i ' nYder 7 Letter 1 Jun 78 Equipment Mail1tenance on AJ lNER S S 1 ----_ 1 Acronyms Apr 80 OH K 1 I AOC News MayB2 Is Tnere An Old Crow In Your Future 1 - - Apptaisal System Nov 78 Letter Appraisal System 'Anon ' Nov 78 Letter Appraisal system I --' Arabic Sep 74 Language in the News Dec 74 Language in the News Apr 75 Language in the News Tetrault E W 1 Africa Nov 77 How Many African Countries Can You Spot 1 1 Afrikaans Sep 74 Language in the News IL - f_-e_-p_- _7 4 J_o_MrTAnalYSiS fl AG-22 IATS Mar 76 Musings About the G-22 IATS Phillips C J May 76 What's Wrong with AG-22 IATS n R --Jun 76 Comments on AG-22 IATS Jun 76 Comments on AG-22 IATS Jun 76 Comments on AG-22 IATS Jun 76 Comments on AG-22 IATS Jun 76 Comments on AG-22 IATS1 Jun 76 Comments on AG-22 IATS Jun 76 Comments on AG-22 IATS Jun 76 Comments on AG-22 IATS Sep 76 More Comments orr the AG-L -2' 2 IA ' T S- r-I---1 I B Dec 74 Lette r c JA l ti le _ I Anon Archives Jun 78 Never Again Gurin J Nov 78 Never on My Watchr I ''' P L r 1 Arms Control Feb 82 Gayler - AAAS 1982 - I1 Artificial Intelligence May 77 Another Controversial Book on Artificial Intelligence 1_- - ---- -- __ Sep 77 The Doctor Is In capable of Diagnosing Chauvenet L R Asia Aug 76 Integrated Analysts for Asia I _ _- 1 w n OCt 76 Another Word on AG-22 IATS IL _ _ IATS illiPS 11 7 The Last Word on C J Jun 77 AG-22 IATS A Vie From the Br idge I ' ' I Assignments Jan 80 Coming Home 1 Sep 77 Another O 7 Letter AG 22 Page Print 'Reader' P L 86-36 May 83 I P L CRYPTOLOG Page lBO 1 4 c P L 86-36 86-36 1 Automation Oct 75 Automation of a TA Process Murphy T Sep D L 771 Arne rieanMag ic A r 8 2 Reyiew The AmericanMa gic I j' Ask Art Apr 79 Letter Ask Art 'Sue' - - Ic J aCID 4009861 SI lSREI'l' ' 'P L Awards Aug 74 Learned Organizations--1974 CLA Essay Contest 1974 CMI Essay Contest CISI Prizes and Honors Spring 74 Jan 76 Lenin and State Prizes Now You See Them Now You Don't I -- Jan 78 Thanks for the Attaboy r- -_-_-_-_-_-_-_-_- Apr 78 Prizewinners in Three Holes Jan 80 CLA News Sydney Jaffe Award Sep 82 Letter Kudos 1 I Blue Russian Nov 7S Golden Oldie Blue Russiand J F 1 -- ------- 1 Budgeting Jul ZBB What Jnthe World Is That W E 1 _ CM News Dec 75 Learned Organizations--CLA Essay Contest CM News Jul 77 CM News What Ever Happened to the CM 1 - - _ 1 Aug 77 CM News What Are They Up to Anyway I 5 Mar T8 CAA N_ews_ Buck Apr May Jun Jul Sep CM News CMNews CM News CM News CM News Oct 78 CM News Nov 78 CM News May 79 CM News Jun 79 CAAN ewS Aug 79 CM News Oct 79 CM News Analysis I Jun 78 Bookbreakers Forum I Nov 78 Linguistics and the Code Reconstructor Buck S H Apr 79 Bookbreakers Forum Apr 80 ooo And In A More Modern Vein LI V AnI' 82 Bookbreakers FCl umOnMachine Aids 1_--1 86-36 86-36 W E S W E S W E S W E S EO 1 4 c P L 86-36 W ' ' 'S' P L I 86-36 1 Conference on Communications ICJ Callsigns Dec 74 An Approach to Callsign Analysis Jackson W J May 78Ca 1 ls igns and WARC-791 jay 78 TA ImPlications of FCC p r o po s a' l- at SIc INT Filby V R Apr 82 Rev ew The Am oo f igati MagigrJ Sep 82 Normandy 1944 L ------ P L P L Calendar Dec 74 A Proposal for Cale @r 1Rl' 4Qr ltl Leahy F T P L 86-36 SeE 781 CAMIlllO Feb 7S c INO News I IM - a-r-- '8 '2-A H ' 'i s t' o r ian I Looks W E S W E S D A Dec 81 Review In The Name of Efficiel cy May 83 -W- E S l' ' I 77I__ I'---- - --JI e 78 78 78 78 78 ' ' M a-r Book Review Feb 75 Review of 'Guide to Russian Technical Tr ansla t ion' by Salemme J 1 Jun 75 The Navajo Code Talkers Dec 75 A Personal Comment on Winterbotham's 'The ULTRA Secret' Tiltman J H May 77 Another Controversial Book on Artificial Intelligence J 1 Jan 78 'Bodyguard of Lie$' Book Review McConnell H Jan 78 'The Man Who Broke Purple' Book Review Filby P W OCt 78 Feeding the Germans Misinformation Book Review Filby P W Nov 81 Review What Do You Think M E o I 1--_--- S H Jan 78 New and Improved Aid for Bookbreaking 1 77 $ep CM News F l -- --- Oct 77 CM News Nov 77 CAl News -T ' '- Dec 77 to1E s ge and News Jan 78 CAANe Wl l ' 1 E S Feb 7aCM News w E c I Hi-II-B-O-O-k-b-r-e-a-k-e-r-S--F-o-r-u-m-' 11 1 C-LINERS Mar 77 Letter C-LINERS Article 'Appalled' Jun 75 Letter Myers Letter ' 'l on ' May 76 Conversation With a Bookbreaker 'Anon ' c J By-Lines Feb 78 By-Lines DOn't Cost--They Pay Bookbreakers Sep 74 Cryptanalysis and Code Recovery Mountjoy M Apr 7S Letter Professionalization of Bookbreakers 1 1 May 75 Letter l Letter 1 May 75 Letter I ILetter _ _ _ Jun 75 Letter Bookbreakers Professionalization of Country Specialists1 I 86-36 Broad Spectrum May 82 Full or Broad Specl ruillt ighting EO 1 4 c -_ P L Career Fields q n 76 How Things Have Changed 1 L Jan '78 ' I'he ChangingF ce of NS ' Anon ' CRYPTOLOG Page 14 SBeREl'i' P L 86-36 86-36 aCID SEeM 4009861 ELL I I Jan 81 Why Do They Leave Nov 82 The Costs of MUddlin g T h r o u g h GO uld R E Career Panels Aug 74 A Short Directory of Career Panels Feb 79 Language Career Panel Clarification of Nov 78 Item I p _ EO 1 4 c CIRC Jan 80 CIRC An Intellig nde D t -B e Celtic Jun 78 Celtic Languages TOday I I __ Censorship Nov 75 When Censorship Backfires 1 __ Central Information Apr 75 A Guide to Central Information C5 Central Research Nov 82 Central Research and The Paper Blob 11 -_ _1 Certification Mar 76 A Comparison of NSA a n d A T A -- Certification Standards LI -- -- 1 Sep 76 Is There Life After Certification' I Changing Environment Jan 80 There's a New World Coming Are You Ready 1 I Chinese Sep 74 Language in the News LI __ Oct 74 Character Building in the people's Republic of China Hamlett G Reed M Oct 74 Even a 5-year-old Child oooo Tetrault ---- --- - -- ---1 E W _ _ _ Jun 76 An Evaluation of a Scientific Chinese Machine Translation I I Jun 76 Notes on Translation from the Chinese I I Nov 77 Backinq into Language I 82 L Acquisitio I----- CIphers Feb 75 Basic patternsoflL - lcodes and May 83 EO 1 4 c P L 86-36 I cipher Systems Oc t 75 comrP u t e r sL' C o m m s a n d L o w - G r a d e__ iphers I __ Catalogs Oct K J 81 Technical Support catalogs 1L - II - Ciphers LI Feb 76 What Is Cipher Text 1 May 76 A Sim pIe ClPher Story- j I Sep 76 Another Cipher by ooo L _ Jun 77 Old Russ ian Manuscr ipt Ciphers Cipher Devices Apr '79 But Mr Boak Did You Ever Try To Get Rid of One in a Hurry D H W CARRIAGE Sep 74 Project CARRIAGE Worldwide HFDF Modernization Plan Webster J B I 86-36 I CISI News A g 74 Learned Organizations--1974 eLA Essay COhtest 1974 CMI Essay Contest CISI Prizes a ndHonors Spring 74 o Dec 74 Lea rned Organizations--CISIForms SpecialI l1 t erest Group on Human Factors I I Jan 75 Learned Orgapizatibns--IA1 News CIS1 News CLA is Ten Year Old CMI News Mar 79 CISI News May 19 CISI News Spring Con erence ' ' ' Cables Classic Classic Classic Classic Classic Mar 19 Apr 79 May 79 Apr 81 ' P L 86-36 Cables Cables Cables Cables Classification May 75 Codeword or COMINT Channels 1 H - - - - Jan 77 Letter Use of Term 'Compartmented' 1 1 J u 1 7 7 C lassifica tion corner I __ Sep 77 Classification Corner XGDS-2 1 1 Oct 77 Classification Corner A Bigger Picture 1 Oct 77 Cl a s s l GBP l c a t l o n c o rner Who Said Kermisch H jiiri19It'g partYTime I ____ I P L CLA News Aug 74 Learned Orgariizations--1974 CLA Essay Contest 1974 CMI Essay Contest CISI Prizes and Honors Spring 74 Jan 75 Learned Organizations--IAI News CISI News CLA is Ten Years Old CMI News Apr 75 Language in the News Tetrault E W ay 75 Learned Organizations--CLA News IAI News LLearned Organizations--CLA News CMI Dec 75 LearnedOr9Cill iza ion --CLA Essay Contest CAA News P L 86-36 CRYPTOLOG Page 15 UlrliB15l3 ViA eSlIHi eltlltftiEbS SUbY 86 aCID 4009861 1 L 8 6-36 Sep 76 Learned Organization$ -1976 CLAEssay Contest 1976 CMI Essay Contest Mar 78 CLA News Sep 78 CLA News Feb 79 CLA News Russian Institute Oct 79 CLA News Jan 80 CLA News Sydney Jaffe Award Jan 81 NCS-CLA Symposium Garofalo C A Comic Strip Sep 77 At Balloni Editor A J S Jan 79 Al BalloQi Editor A J S COMIN'l' Sr-e pc 7 4 lC O M I N T Analysis oit '--- ---- 1 - -1 L I_ _ Clustering Apr 78 A Little PEP Talk IL- _ _ CMI News Aug 74 Learned Organizatiions--1974 CLA Essay Contest 1974 CMI Essay Contest CISI Prizes and Honors Spring 74 Jan 75 Learned Organizations--IAI News CISI News CLA is Ten Year$ Old CMI News Jul 75 Learned Organizations--CLA News CMI News Sep 76 Learned Organizations--1976 CLA Essay Contest 1976 CMI Essay Contest Jan 79 CMI News Aug 79 CMI News I --------EO 1 4 c Codes P L 86-36 Sep 74 Cryptanalysi and Code Recovery Mountjoy M Feb 75 Basic Patterns ofl Icodes and Ciphers I 1 -_ _- ' Jun 75 The Navajo Code Talkers Jun 76 The Marquis and the Medium Nov 78 Linguistics and the Code L Reconstructor Buck S H Mar 821L - 1LI ----I EO 1 4 c Collect ion F L 86- 3 6 Aug 74 What Isa Collector 1 Dec 74 The Newi CoUection Crl 't-e-r-l' a-' Ir- L - Aug 75 Abdul and His 40 Tanks Mason F O Feb 76 The 1 Isystem 1 1 EO 1 4 c P L 86-36 Jan 75 The case for COMINT Readers H G Apr 76 Golden Oldie Hidden Losses in COMINT Production GouldR E May 76 COMINT in the Russian Navy WWII Jan 78 COMINT COMSEC and Hilbert's Tenth I I Feb 79 The COMINT Chain 'Gang ' 1 __ 1 4 c 86-36 ' 'P L 86-36 Communications Changes Apr 75 A Carom Change at Rama$unStation 1 1 Oct 82 Golden Oldie The Realitypf Communications Changes I I Computers Aug 74 Right-to-Left Text Sorts Are Not Impossible Dec 74 Flag-Wavlng programmer John G Jan 75 The Yawn of the Computer Age or When Your Terminal is Terminal I Jun 75 professionalizing in Computer Systems I I I I M c t 1 Pr inCi leS of Icovermand 86-36 Jun 76 Th1 jFollection System Murphy T Aug 76 NSA Cryptologic COllection r-_ _- OCt 76 Another Word on AG-22 IATS IL-_ _- I M A Jan 77 Where Were We Mason F O Nov 77 Director's Memorandum r - - - - - - - Guidance' Inman B R Nov 77 Objective Satisfaction Score Collection performance ' Jan 78 An Early NSA Proposal for Satellite Remoting 1 IN lte W M Jan 78 What Ever Happened to COPES IL- -I W E Feb 78 Collection-Support TA is Notfo Every o 'n 'e' ' I - I - - - - - Jun 7 I Jan 8''f '' K I ' T T 'IW ' AK - E - rl I- An Overview J I of PrOjectl 1 '----- Contemplating computing LI_ _ _ _- I1 Universes Galaxie$ Stars Bars and Concepts Phillips C J Ear ly Days in NSA Computing iChauvenet Colo Code Sep 74 Gary's Colors Garofalo C A Feb 82 Golden Oldie Simplicity in Color '1'6 P L EO 1 4 c P L 86-36 May 83 86-36 CRYPTOLOG Page 16 I1tmf5 0 HI eSlHII'l' efW1U8 S SUM' EO 1 4 c am 86--4009861 i L' Jan 80 Analysts of NSA Arise J Nov 81 How to Create A user unfr l e n d l y ---' System 1 Nov 81 The PPC Is Coming 1 Dec 81 In Pursuit of Faster Horses Younget Women Older Whiskey and More Money 1_ 1 D L Dec 81 PLATFORM How Did You Say That WQrkS i 1 EO 1 4 c P L 86-36 E HilElRB'I' 1 Feb 82 Human Factors Corner Some Advice to Users of Unfr iendly System 1 Feb 82 Software - AAAS 1982 1 - 1 Mar 82 A Personal Computer A Current Cryptanalysis Support '1'001 1 Mar 82 Human Factors corner c o-n-s-u-m-e r ' 'v-s Computer A Review I Apr 82 Personal Computer Application t 1 86-36 Jul 7'7 Whi h Tape Has the Intelligence project _ IIGUrin J CRITIC Nov 78 Wedding Bells and'1'hat Old Gang of Mine Sawyer E L o CRT Mar 78 A Linguist Looks at the TUbefi R S --Mar 78 The Hand Is Not Quicker Than the Eye 1 1 Aug 82 Human Factors Corner Video Display Terminals and Vision of workers i M E ---- I Cryptanalysis Sep 74 Cryptanalysis and Code Recovery Mountjoy M Oct 74 An October Overlctp l - -_- -_ 11 Nov 74 Answer to An October OverJa c J 1 1 J E_ Oct 82 What's The c ooo pass Word 1__- 1 u Dec 82 Going On-Line With Information 1 id8 Nov 74 New Trends in the Teaching of GBPryptanalysiSil Gur in J ip GBP 8 60 36 Dec 74 Puzzle Secret Messages 'Military Computer Aids Cryt tCinalytics ' Apr 76 Computer-Aided Transcription of May 75 TA Jlal' dlllaiden ofCA MasonF - O _ _ Jul 75 Re-psychlingt e Code Clerk 1 1 Jul 75 Too Many Garblesl _ _ '- I Computer Security Aug 75 Twenty Years of Transpositioii c J J E Mar 79 Computer Operating System jUg 75 Ty ewriter Random -- A New Look YU 1nerabilitiesd 1 Jun '82 'Some 'Reflections 0n The 'Reality of Oct 75 Letter Typewriter Keyboard C6l1lPuterSecurity HanyokR J Oct 82 What's The Good passlword I 1 Oct 751 F '----- - - - - - - - - - - - Dec 82 passwordsd I Apr 82 Word processirtgr I n A 4 I -r Sep 82 TSS Revolution I_ _ _ 1 Oct 82 Human Factors Corner Text Editors I 1'---- 86 36 1 u M F ea yb 76 wA h Sati m pIlsec ciPl phhere r TSeXt 'otr y J P L S6o 3 76 Computer TA ---- '- ----' Aug 76'''TheHierarchical Clustering of Apr 80 A Traffic Analyst Looks at Computers Cryptanalytic Data 1 1 Nov 76 Teacher Learns a Lesl30rt I COMSEe want to Play with a Pickfai are Jun 75 COMSEC Familiarization Do You Need It Jan78 TheMal1 Who Broke Purple' Book Jan 78 'Bodyguard of Lies' Book Review I I d r-I---- I P' L Jan 78 COMINT COMSEe and Hilbert's Tenth 1 1 Apr 79 COMSEC SIGINT Rela ionS Boak o G Nov 82 COMSEC Challenges 1 EO 1 4 c COPES P L 86-36 Nov 76 Check Your Morse Front-End Allgnment 'Anon ' Jan 78 What Ever Happened to COPES W E 1 '---- Jun 78 Letter I J1 Article Gilbertson E A Sep 78 Le tter Gilbertson Letter LI Coverterms Mr75Coverteritls Filby V R f i ' P L 86-36 8 vi w Fi LbyP W Apt 78r o k i ngatJ1r 1 - ''---_ _ J I saacL ' May '18 The Fu ture of Cryptanalysis Lutwiniak w JUn7S c J L S '----- -- -- ----' Jan 82 Letter Cryptpl'lalysis Articlei A Mar 821- ----- - - -- -- 1__ __--- ' ' 1 peg 82 pger Anybody Here Remember PURPLE 1 Cryptanalysts Aug 74 What Should You Expect or The Analysis ofCryptanGiJ ysts uuC r--- lI _ Jun 76 How Things HaveChangedHI _ EO 1 4 c P L 86-36 May 83 CRYPTOLOG Page 17 4009861 Crypto-TA Jun 75 More on Squaring the Page A Crypto-TA Function Mason F O Aor 79 A Somewhat Larger Problem 1 W E ------ Aug 82 An Old Problem I------ l _ Oct 82 Answer An Old Problem IL - 86-36 Data Flow Aprea Data Flmi -Challenge of the ___ Cryptograms Sep 74 Secrets of the Cryptograms 1 Altars -The Callimahos L D EO 1 4 c P L 86-36 Cryptologic Mission Feb 79 Let's Not Forget Our Cryptologic Mission Gurin J Apr 79 Letter Gurin Article Feb 79 l On Ve r a ' l M R c 8 6 '36 CSI News Dec 75 Establishment of CSI Newsletter 2--r-- Degarblr-in_g sep77I D L Dec 77 Letter c JArticledL __ DESKPAD Nov 75 DESKPAD A prbgrammers's Tool 1 Cyrillic Jun 76 Transliteration or Cyrillic IL G Bl ob May 83 86-36 --- H 1 Data Bases Jan 75 How Clean Does a Data Base Need to Be LI--- --- 1 Feb 79 Literature Search On-Line Miller M E Jan 80 CIRC An Intelligence Data Baf e P L 86Deception Feb 76 How Do We Know It's Trlie Filby V R May 76 Some Pr nciples of Cover and J ci t t I and Thepa r I J lJL79I EO 1 4 Nov 82 Central Research 86- I I 1 Crypto Devices Aug 82 AFCEA 82 and ICC-82 New Cr pttl ti L Dates Jul 77 Dating Game Williams D H Aug 77 P9 script to 'Dating Game' 1 Cryptosystems Oct 75 The Do Xa Pads Wiley E Sep 78 1 Data Standards Nov 74 Data and Definitions Calling Things by Their Rightful Names I Nov 78 Data Standards Center Pattie M T Dec 78 a k a Sam 1 I Feb 79 Data Standards Without Tea s I_ _- I P Q Feb 79 Well Maybe a Sniffle or Two ooo Pattie M T Mar 79 Data Standards Without Tears A 1 Commentd Mar 79 Letter Data Standards LI Jun 79 ooo Mr Pattie Replies Pattie M T Jun 79 Data Standards Without Teetl l IL' -__- I P Q Aug 79 An August Baudy 1 1 Jan 82 Data Field Nami g COding Conventions at NSA I I Aug 82 All I Ever Wanted To Know About DES I Cryptology Jan 76 Football and Cryptology I Feb 78 First Lady of Navy CryptologytL R P D vices 1 c J Data Sorting Aug 74 Right-to-Left Text Sorts Are Not Impossible 1 I CRYPTOLOO Aug 74 CRYPTOLOO A Letter of Introduction Wolff H E Dec 74 CRYPTOLOO Index for 1974 Jan 75 CRYPTOLOG Index for 1974 Dec 75 CRYPTOLOO Index for 1974-1975 Nov 76 Letter Cumulative Index Bostick C W Dec 77 CRYPTOLOO Index for 1977 Mar 78 An Idea for an Article Miller D E Dec 78 CRYPTOLOG Index for 1978 Jan 79 Letter From the Publisher Lutwiniak W Mar 79 Readers' Survey D H W May 82 CRYPTOLOG Number ing 1 Data Security Oct 78 Data Secur ity and Human Error J M Moustier 1 Cryptographic Apr 75 Psyching the Code Clerk L 1 80s PhillipsC J DIALOO Aug 77 DIALOG Avai able at NSA Feb 79 Literature Search On-Line Miller M E Dictionar ies Feb 75 Glossaries versus Dictionaries Which Should It Be Gurin J Feb 75 The Devil's Dictionary CRYPTOLOG 813EURM' ' Page 18 DOCID 4009861 Op o Aug 76 Note on FRANCOPHONEGLOS May 78 Project UTENSIL The DDD Data Dictionary Directory I S Jun 79 io Are Your Stamina J 'airbanks s Jul 79 Wilt Thou Angelin oo Fairbanks S I Acq isitlon Document Review I Economic Intelligence Oct 81 Economic Intelligence Problems and Prospects I I Editor's Note May 76 Scraps from the Editor's Desk Salemme A J Feb79c J and a Little Bit of Luck Williams D H J i i ---------1 P L 86-36 EW dun 75 The Role of the Electronic Warfare Advisory Element JEMAE of N$A I I Oct 76 Russian SIGINT and Electronic Warfare Efficiency Dec 81 Review In The NamE of Efficfency I 1 EL NT Nov 74 G A What Where Why c J t sB L cL II e- -I-- IL- I I Language in the News Where Does 'Does' Come From Tetrault Language in the News The Uses of Elegant EngliSh 1 L -_ _ 77 Choose Yel 77 Plain Engli s ' h- 'J---L ----- 77 Vich sS R-r' ' igJ lt A J S 77 Whac EverDoes 'However' Mean May 83 '--_ J Jan Apr Jun Jan 76 76 76 77 Leo in October I Murphy A o I o One Day in Danang 1 Letter 1 IArticle Where Were We Mason F O Forms Sep 77 Speaking of Logging ooo I__ __ FRANCOPHONEG LOS Aug 76 Note on FRANCOPHONEGLOS P L 86-36 French Aug 74 Nice Busman's Holiday for One NSA Employee I Sep 74 Language in the News I I 86-36 -'- 'EO 1 4 c P L 86-36 A Jan 78 The JoYs and Frustrations of PluralDropplng A J S Mar 19 Fairbanks on English Fairbanks S Apr 79 More Fairbanks on English Fairbanks P L ---' Field Stations Apr 75 A Comm Change at Ramasun Station Reese J C Oct 75 The Danang Processing Center 1 I I _ _ Jul 79 EXPERT 1 Emitter Identification May 79 Emitter Identification Techniques May May Jun Nov EXPERT Jul 79 EXPERTd Schlau - c 'l h lR Hr -------- I Jun 76 Computers In The ELI NT and Telemetry Business i-'------ I Aug 76 Yes Don ThereisariELINT c J W L Nov 76 Letter From c nadad _ English Dec 74 Jun 75 E W Aug 75 Nov 76 J R I Exploi ta t i o n Feb 781 AQr 76 WilL the Real ELINT Please Stand Up I' Equivalency Tests Nov 75 NCS Offers Course-Equivalency Tests Apr 76 NOS Offers Course-Equivalency Tests Clar if ication Evaluation Oct 77 Kl SeA Field Manalement and Editorial Comment Feb 78 No winnie You've Got It Upside Down Too A J S Jun 82 Letter Editorial comment 1 P A ------ I li quipment Jun 78 Equipment Maintenance on NER Bernard R Oct 78 Will It Really Do the Job 1 Apr 79 But Mr Boak Oid You Ever Try To Get Rid of One in a Hurry D H W Oct 79 Another Source I I Oct 79 Snowballs On The Roof Filby V R Aug 82 AFCEA 82 and ICC-82 New Crypto Devices 1 t I Jun 82 Human Factors Responsible Documentation I I EOL4 C J P L 86-36 8 6- 3 6 S MCI y 79 More 'Fairbanks On English Fairbanks Documentation Se 77 Is There a Docf or in the House Jan 80 System L Aug 76 Language in the News 'EO 1 4 c P L 86-36 CRYPTOLOG S13EURfS3'f Page 19 IIldHHd3 VIA 881HIl'l' 8IIUJUI3I S ElUI Y 4009861 ' P L French Connection Dec 76 The French fr ied Connection uGinothe Genie I I HF Frequencies Jan 79 The Return to HF I I Jan 82 HF - The Rebirth I Jun 82 Amateur Spread spectrum I HF F Sep 74 Project CARRIAGE Worldwide HFDF Modernization Plan Webster J B Nov 74 1 I What Where Why G A EO 1 4 c P L 86-36 Hiring t74 News from NCS--Agency Resumes Hiring Feb 76 The I Oct 77 Update onl Isysteml_' J Games Oct 76 A Vexing gency-Wide Problem Mason F O Nov 76 Answer tb 'Vexing ooo Problem' Mason EO 1 4 c F O P L 86-36 Geography Dec 74 Maps in Mind--A Photoessay I __ Dec 74 Puzzle Citizens of theWo Ld Nov 77 How Many African Countries Can You SPOt 1 Apr Oct Dec Feb Apr 78 78 78 79 80 Ge _ _ _ _ _ Conoces la Know Your Bien Geography How Do You Spell Pe' Know Your Geography Geographic Trivia R tur tOHE' 1 Jan 79 Th e o n o o Jan 82 HF-The Rebirth '--- 86- 3 6 J c J i t i f r coursein 'SIGINT Dec 75 Linguists From The Meltillg Pot Gould R E 86-36 History Feb 75 The GulfofT6fikin Inci W O lay 76 COM in auSsian Navy WWII Sel 78 Soviet COMINT and the Civil War 1 I German Sep 76 Language in the News Nov 77 Backing into Language Acquisition 11 --_ _----' P L 86-36 Glossaries Aug 74 The New Traffic Analysis Glossary Oct 74 Golden Oldie An U official Glossary of Weasel Words Feb 75 Glossaries versus Dictionaries Which Should It Be Gurin J Feb 76 Expletives Deleted Glossing Over a Glossary I Aug 77 Expletives Deleted Salemme A J '-------'1 Jun 76 O EO 1 4 c Thel 'l------ K oltec lo y fem 76 Letter 1 Murphy IArticlel __ Graphics Dec 75 Graphic Analysis of Linear Recursive sequences I 1 Dec 76 Graphic Names I Human Factors Jun 77 Human Factors and s stems Design An Estranged RelationShiP r_ Oct 77 Human Factors and l e use of Microfiche Readers at NSA Snow D Jan 79 Human Factors Newsletter Jan 82 Human Factors Corner Information I system Feb 82 Human Factors With My Papers I Feb 82 Human Factors Users of Unfriendly Mar 82 Human Factors I Corner But What Do I Do I Corner Some Advice to System 1 1 Corner Consumer vs Computer A Review 1 Data G a1t her 1' ng o o May 82 Human Factors Corner How Do We Spend Our Day I ---- 1 Aug 82 Human Factors Corner Video DisplaY Terminals and Vision of workersll __ M E Oct 82 Human Factors Corner Text Editorsi I --l' -o - II Nov 82 Human FaCt r Corner How Do peopl Qrganize cooperativ e Work I------ --- - i Dec821 a swords 1 ____ GUPPY Feb 75 Replacement of the GUPPY Library I_ _ I Hebrew Sep 74 Language in the News I IAI News J HELIPAD May 79 Project HELIPAD An Epitaph '---_ _- J M L May 83 Learned Orgariizati9ns-- 1M NewS ctSI News CLAiI Ten ars Oldl ' ' I NewlOl May 75 Learned Organizations -- LA News t News P L Jan 75 CRYPTOLOG Page 20 IIA I5I l un QQUHl'i QHJldlUI lI5S eUl JPY 86-36 aCID 4009861 P L 86-36 J'apanese Nov 76 Tgujou s La Politesse Japanese Buck Indicators Dec 82 Development and Correlation of Indicators 11 11 S H Jul71 The Transcription Skill Concepts and Teaching Meth9dologies I I Jan 79 W W IfJapanese Translation at Ar lingtPri Hall Station Information Jan 82 Human Factors Corner Information systemlL 1 May 82 The NSA Information Desk 'No comment' -- I JPRS Be INKSTAND Jun 77 SIGINT Welcomes INKSTAND I'-- I I 77 JPRS Language Reference Aids I Keyword Spotting Mar 77 Letter Keyword Spotting Gurin J EO 1 4 c Intelligence Dec 75 What Are We Abo u t _ Figments or What Apr 7 6LE tter comlllent -s on WhgLAreWe APout _ - -_ 76 Letter What Are K I r I 1 rIWAKI II - --r----- 1 1 P L 8 6 - 3 6 I __ -_ _- - - - - - - - Intelligertce CQm u ity Aug 77 New Directions fortl1ey S Intelligence CommunitY1 Gaddy O W Intern Program Sep 74 A Long Hard Look at the Intern Program--Program Philosophy Recruitment Part One 'Exinterne' Oct 74 A Long Hard Look at the Intern program--Selection and Orientation Part TWO 'Exinterne' Nov 74 A Long Hard Look at the Intern Program--Motivation and Morale Part Three 'Exinterne' Dec 74 A Long Hard Look at the Intern Program--What Happens to the Graduate Part Four 1 'Exinterne' Feb 75 Letter Exinterne Articles Tetrault E W _ Apr 75 Letter Exinterne Articles 11 Interpreters Feb 75 The Faithful Echo--The Role of the State Department Interpreter 1 1 Oct 77 The Perils of Being a State Department Interpreter Feb 78 The Unseen Go-Between Gurin J IRA Aug 77 What Is an Information Research Analyst 1 I EJ Oct nlL - I I IL ISHTAR Jan 75 The SIGINT Users' Handbook or What's an tSRTAR 1 I Language Aug 74 Nice Busman's Holiday for One NSA Employee Dudley B Aug 74 The Language of Beisbol in Everyday Talk I santiago-Ortiz R A Sep 74 Language in the NeW13 L 1 Oct 74 Character Building in the People's Republic of China Oct 74 Even a 5-year-old Child ooo '1 'etrault E W P L 86-36 Nov 74 Purity of the Russian Language- '7 f lavophiles vs westernizers Dec 74 Language in the News Decm74mThemOl ection Jan 75 The Case for COMINT Readersr H G Feb 75 The Faithful Echo--The Role of the State De artment Interpreter I 1 Apr 75 J anguage in the News Tetrault E W Jun 75 The Navajo Code Talkers Jun 75 Where Does 'Does' come l Qm 'j' ir au5 _36 E W lAm 15 A Filfor t heLanguage Problem I I I Aug 75 Language in the News Oct 75 Language Lessons Learned A Personal Memoir 1 1 Oct 75 Tactical Language Exploitation A Lesson Learned Nov 75 Golden Oldie Blue Russian11 J F --Feb 76 Expletives Del eted Glossing Over a Glossary I Mar 76 1976 Language Meetinqs and Conf erences Mar 76 A Comparison of NSA and ATA Certification StandardsrlL - I II IRONHORSE Oct 75 IRONHORSE A Tactical SIGINT System Oct 76 Letter IRONHORSE Article KRYPTOS News Feb 82KRYPTOS A NewSocletyd Dec a2 KRYPTOS News I I 1 May 83 CRYPTOLOG Page 21 seeM' ' P L 86-36 Ih IffJfaH VIi' e8 JIU'f EURnt UUE S eUEiY DOCl qOl 4 -9 Q B 61 P L BQ-36 r i i C Pt10 r' I 1 y mi i d 'th 86-36 Opera t 9nsr J I ' i Feb 19 Lal guage Caree f' aneh ClatJf cilti iQP Apr 76 How Do Adultsr earnLanguage 1 1 r L i dt l n u eL g 1 v nIi m i I I i 1 Learning Buckley D May 79 Run T' I S T ht ough tcl1 1r T iansdript 6n May 76 What Language Problem PattieM T Machined ' V i Jun 76 An Evaluation of a Scientific Chinese May 79 Where DQ G d ' 1'r anscr fqers coDie iF ot 1 Machine Translationl' ' 1 1 1 ' --i--i--i--i-i Jun 76 Notes on Transl'cifion from the Chinese Jun 79 Letter 'Sixth Lanw1a Je' LJ_ -i- ul l he lug 9 signtan uagel i I 76 Language News Aug 76 Research in Soeech Perception Dr Jun J9 The 'Baltic' Encod rs' I--i-- - j Ruth Day Tetrault E W JU 1 79 Ge ars 'of the Motit1 l' I 1 Sep 76 Language in the News Jcil 79 Letter' 1 I Article Apr 79 Oct 76 Language Skill F i l e l l Mooney K Jul 79 wilt Thou Angeiina j Fairbanks S Oct 76 Some Ideas about Mechanized Language Ati 9 79 Language Proficiency Certificates for Working Aidsll Military personnel Nov 76 Golden Oldie The ThIngs They Say 09t 79 NCS Summer Language Program o Miller D E Apr 80 LIME-A OHIO LEEM Peru Salemllle Nov 76 The Uses of Elegant English 1 A J J R Apr BO P16 Langul e and c ryptol09iC Librhry Nov 76 Toujours La Politesse J panese Buck S H J a-n 8 i Wh y Do T hey Leave LI o _ _ Dec 76 What's In a Name 1 ' I Aor 81 Grading The Russ ian PQE LI J1 Jan 77 Letter rticlel Kenny Jan B2 Letter 1 rArticle Chauvenet M M L R Apr 77 'It's Got to Get Out Todayl' Blank F Feb B2 Native Scripting of Languages Jul 771Ji Qmu B c I I S H A pr 82 Partial Machine Translati6n Final Jul 77 Tool Lanquages I IP'mLu' 8 6- - 3Eiu mmReport uuprattmo L lm u luuuu uuuuuuuuuu Aug 77 EXPletives Deleted Salemme A J Nov u'l'he c stsu of uMuddliriif Thro gh Gould P L Sep F ICE uuuuuEO 1 4 c D L P L 86-36 Sep 77 JPRS Language Reference Aids Language Identification I Juuuuuuuuuuuuuuuuuuuuuuuu Jan 80 LIP IL II Sep 77 The c Jsemi narJ'I99 r cllll lu Ihh Wo L L e a d e r s h I p ' I ' ' I 1 I p I 771 fV 77 Back ing inj m m isitionltt82 Leadershif' 'h ' l roMY P L l Nov 77 Lanquage processinl uXC t ll lll luumuuuuumumuuuuuuu 1 Dec 77 uul Lexicggraphy' ' fo IL t Feb 75G16s aries versus Dictionaries Which r i ln 8 aboY I _ _ _ ---IL Feb 78 The Unseen Go-Between Gurin J May 78 Uncle-a Sam Wantsa You d Jun 78 Celtic Languages Today 78 Is A Translr a L r I PrOfesSional ' ' t P 'I n ' ' l ' c I 86-36 1 J n r9 Letter Libra' Y Changes I I Apr 80 P16 Language arid Cryptologic LibrarYl I s Y e ' lH vaJPY$h l I' 1 'Mn g u is iCS'' I Innnnnn Dec 7B Agency Summer Lang'ua je'Study Iillckley D Eo Gud J Library P L Feb 75 Replacement of the GUPPY Library 1 l Sep 7B Letter Pattie Art ic le Buckley D See 78 Soviet COMINT and th Civil War I r 1 $ ' 4 hs llle U9hts6nL icographY Buck What Made Them Do I t Langllage Self- 1 al til 8 6- 3 6 oo Nov 78 Linguistica andth Cod ReCOns t c tO BUCk s M ' ' ' ii May 83 CRYPTOLOG 8SeRi3'f Page 22 P L 86-36 IIII11B 123 VIA 8911111'1' eUl1l1UB158 8U Y 86- 3 6 aCID 4009861 SBEURRl3'P 'f ' L Linguists Apr 75 Machine Course for Linguists Tetrault E W May 75 Are We Wasting Linguistic Time c J M R Aug 75 Linguists -- We Need An' Experts Yellow pages' 1 I OCt 75 Linguists --- YO U H ave an Expert to Call I I Dec 75 Linguists From The Melting Pot Gould R E Sep 76 Machine-Produced Aids for the Linguist Part I Salemme A J OCt 76 Machine-Produced Aids for the Linquist Part II Salemme A J Dec 76 Let's Give the Linguists a Bigger Piece of the Pie d I Jan 77 What If the Linguists Disappeared Myers L S Mar 77 A Few Thoughts on the NSA Linguist 'Anon ' Mar 77 Letter Butcher Article Buckley D Apr 77 Letter Buckley Letter Tetrault E W Jun 77 Letter Buckley Lette Sep 77 Letter Lebanik Letter Buckley D Mar 78 A Linguist Looks at the TubeLI R S Jun 78 As I Was Saying Two Years Ago ooo Pattie M T Jun 79 Teaching Computer Science To Linguistsd I Aug 79 On Coming of Age at NSA Confessions 6 36'of n Ex'--Lin9uist 1 I Qct 79 Letter c J Article Aug 79l' Taylor I f L Jan 80 LJ Rd -- __ _1 Jan 80 NSA CSS ilitary Linguist Program Apr 80 Help want eo 1 I Auq 82 Linguist Machine LI LIP Jan 80 LIP Machine I lHg nce o Jul 5Macl1ine Intelligence--PrQm se or Delusion 1 I Machine Translation Jun 76 An Evaluation of a Scienl ifipchinese Mach ine Tr an sl a tio rdL J Management Aug 74 Golden Oldie The Manageroent SUrvey of the Philharmonic Sep 77 Knowledge Resource Management at NSA I I Mar 78 A' Donkey in Your WH A T Jul 78 The Bucky Ba1ance I __ Management Directives MaY 75 Hooray for PMDS I _ _ -- Managers Feb 78 Senior Military CryptolQgic Supervisors Course I Apr 78 Accentuate the Negative 1 Maps Dec 74 Maps in Mind--A Photoessay L _ _ - IL MAROON SHIELD Nov 77 Director's Memorandum Guidance' Inman B R EO 1 4 c P L 86-36 Mathematics 78 Tell Me I'm rust a Sinobibliophobe I - lUI Mechanized Translat i o n Jul 75 RAPIDT _ Ie I Meetings Mar 82 Rules For The Camel Corps 1 C 1 LI Lycian Sep 74 Language in the News '------ Memorable Memo Aug 79 Memorable Memos Lithuanian Jun 79 The Baltic Encoders Reiskis A - I LI Machine Apr 82 Partial Machine Translation Final Reportd'-I I Messages Feb 82 A Time for Change IL_ - IL Jun 82 Golden Oldie Repotting Message vo1umes LI J Meteorburst Mar 82 Meteorburst Commurtications I Machine Aids Apr 75 Machine Course for Linguists Tetrault E W Sep 76 Machine-Produced Aids for the Linguist Part I Salemme A J Oct 76 Machine-Produced Aids for the Linguist Part II Salemme A J OCt 76 Some Ideas about Mechanized Language Working Aids I I Apr 82 Bookbreakers Forum On Machine Aids I 86-36 Microcomputers May 82 The Micro Revolution Arthur Young Study-CommentJ J Micrographics Sep 76 The13ib1e and th Was 'iil tqton Monument o i Snow D Oct 77HumanFactors arid the Use of Microf iche Readers at NS Snow D Dec 77 Conversation WHh ia Micrographics Pionee r I I May 83 _ CRYPTOLOG P L Page 23 IINIQ15i1 HI 86-36 QQrtIN9 8I11 1'RJl3158 Qll15JPY aCID 4009861 P L I OCt 79 Snowballsc Tne Roof Filby V l pr 80 What To Db About 'FANX' IL - May 82 'I'l leNSA Information Desk 'No S l1llment ' I I Oct 82 Not Secret Anymored Oct 78 Back to Square one 1 Oct 78 Reduction Ratios in Micrographics Snow D Military Jul 78 The Soviet General Staff '---_ _ 8 6- 3 6 I ---- - I Minicomputers Jun 78 A Computer Scratch Pad at Home or at Work L I 1 Jun 78 Minnie's Mini Kenny M M EO 1 4 c Missilers p 86- 3 6 Nov 771 J 1 L ----I ------rJ__ Molecule Superseries Feb 75 Golden Oldie Establisl1lt@ri'e o f c Molecule Supersedes P L 86-36 On-Line Aids Dec 82 Going On-Line With Information Aids1 Gudn J On-Line Data Aug 77 DIALOG Available at NSA Feb 79 Literature Search on-Linell OPELINT Nov 81 OPELINT Is Alive and Well In B Group I f F L Order of Battle u Letter Rosenbluh K E MORETOWN Mar 78r-1-------1JI' --- Jul 78 Le-t-t-e-r- 1-- -011 r ticletl _ Artt r---- P org gizations Dec 78 'l'0WhQIl1J t 1a Concern Grant L G Feb 79 Letter GranLArticle Ql 78 Pattie MOSES Jun 79 The Story of MOSES Gardner D G M o T Apr 81 Middle-Age Spread MSI Dec 75 The Great Soviet Shipbuilding Mystery Williams D H Multichannel Oct 78r 1------------ 11 __ s' ' ------- ID M i l i a t e T e c h n i g y e I Organize Work Nov 82 H u m an Fa tC'rs e or ner HQw Organize Cooperative work 1 Do People ---' Overheard Oct 78 Overheard in the Burnbag Line Api 82 Overhea d While Standing in the Burnbag Line D A EO 1 4 c Navy P L 86-36 Feb 78 Soviet Navy Command Post System Pattern Recognition Dec 77 What Is Pattern RecognitiOn r J D A 1_ _----1 NCS Oct 74 News from NCS- Agency Resumes Hiring of LICs NCS Offers Cour se in 'SIGINT Appreciation' Apr 75 Language in the News Tetrault E W Nov 75 NCS Offers Course-Equivalency Tests Apr 76 NCS Offers Course-Equivalency Tests Clarification Oct 79 NCS Summer Language Program Jan 81 NCS-CLA Symposium PERCIVAL Nov 76 Check Your Morse Front-End Alignrnent 'Anon ' Performance Feb 82 The Internal Performance Evaluation Friend or Foe 1 1 I Personalities Jul 71 Match Them Up Jan 79 GOlden Oldie on F i r $ t-O p-e-n ing Kenney's 'Statistics' Mountjoy M oJan79 Henry CEURment and other Phantoms of the Ope ra ticms 1 1 Oct The Roads Around Us Chauvenet L R I Network Aug 75 Computer Network Resources In C5 NSA May 76 A Soviet View of NSA Dec 78 A VIP TOur through the A t t i c o f N S OCI I ' ' '1 PEP Apr 78 A Little PEP Talk I Performance Appraisals Jan 80 Between The Lines of your Performance Appraisal 1 Negative Intelligence Oct 78 A Method for Measuring Negative Intelligence I I I I FI P L May 83 86-36 86-36 CRYPTOLOG Page 24 IlUi8Wl VliI QQUIUIf elIMlUBh8 atlh' ' 'p L aCID 4009861 L Personality Sep 77 Tribute to the Guru 'Class 32' Dec 82 Questions In Search of a PQE 'Schmedlapp' Personal Computers Mar 82 A Personal Computet A Current Cryptanalysis Suppor tTQol _ _ t Apr 82 Personal Computet Application Predictions Feb 76 The Prebendary and the prophet 1 Dec 76 But It Looks Like the Real Thing '- ' ' I Mar 77 Letter 1 H L Personnel Feb 78 First Lady of Navy Cryptol y -H- 4 i l R P Jan 79 How do You Tell These Two Clowns Apart i Plaintext May 82 Letter Plaintext I __ __ f PLATFORM J Mar r' 77 An overvie w of proJ ec 1 '1 ____ Professionalization Dec 74 The New Collection Criteria I _ __ Apr 75 Letter ProfessiOnalization Of Bookbreakers Jun 75 Letter Bookbreakers Professionalization of country Specialists - 'H PMT Apr 82 Partial Machine Translation Final Report f A 'u-g 7 '7 L e t t er I---i---- ----'1 Ar t i c Ie 1 ' 'Weeson Portuguese Sep 74 Language in the Newsd I Jan 78 Thanks for the Attabo 'y - '---------- I PQE Jan 75 The Ca$e for COMINTReaders H G Dec 76 Why Can't They Design a Good SR Test Bjorklund K Apr 77 Some Thoughts on the Russian PQE I_ __ I' Jun 77 Letter Clark Article 1 I Feb 79 In Defense of The Inde fe-n-s ib le- Notes on the Russian PQE Tetrault E W _ Apr 81 Grading The Russian PQE I Jan 82 Letter IArticle C l 'h' 'a - uv 'e - n - e - t --L R May 83 T sting and What To Do About 1 for Promotions Nov 77 A PiQPOsgd Cytgfor the Time in Grade Syndrome L -II Mar 78 A Proposed Cute ror the 'PerfOrmance Syndrome' 1 Dec 78 Some Tips onGe t j ng Promot ec 'l ' V C Feb 79 Letter I tticle o Dec 78 Bostick C W Aug 79 But On The Other Hanc 'l Aug 79 NSA Promotion B6Atrls How They Work Poems Jan 77 Executive Orderi 11652 'Leiner' Jan 81 Some Things Never Change Feb 82 A Wail A Complaint ana a Melange Snyder S S 1 Proficiency Aug 79 Languagepr fir c i eo n lc ly c eo r t -i ficates Military perso rel 1 I May 75 Hooray for PMOS _ _ ---1 I 75 Professionalizing in Computer Systems Apr 79 Fear Of It 1 PMD Polyhedrons May 77 The Polyhedral War _ Apr 77 Some Thoughts on the RussianPQE I Oct 78 ContinUil'lr I rofessionalization PL1 Nov 78 Formatting PL 1 Source code J I IInterview 1 Production Apr 76 Golden Oldie Hidden Losses in COMINT Production ould R E IJun Dec 81 PLATFORM How Did You Say That Works K J 1 -_---' R i Jun 82 Letter Personal computer Article 86-36 1 Jan 81 How to Improve You romQtipn Potential ' Jun 82 Who Wants A P omot dri Anyway lug 92 what prlmotioniBoaras want 11 '-'-' ------ Publications Jan 75 The SIGINT Users 'Han qbool 9 What ' s an ISHTAR L May 76 About the NSA SWINT suipmary HuntW Sep 76 Foreign Publications PrQcurement Program ' 1 Mar 77 Revised Technical SIG N'i Marlual in Preparation Filby V R o o ' May 78 Project UTENSIL he DDQQa# Dictionary Directory ' Dec 78 'NO No Nanette ' Means 'JPY ' A J 5 Jan 79 Human Factors Newsletter CRYPTOLOG S SSHl3'l' 0 P L 86-36 Page 25 I1i UB d3 'Hi E 8 1fU'l' E I1i tfti'BflS 8tiflJPY ero 4009861 Sl3eIll3'f 86-36 Jul 79 Russian Handbook of Spoken Usage Vol 3 Jan 80 System AcquisitionDOCl 1ment Review I Mar 82 Old Phone Books Never Die Nolte W M Punctuation Nov 74 The Apostrophe SomE Thought's Oct 76 The Winnah Kig ApOstrophe Aug 79 Punctuation More Than Meets The Eye 1 1 PURPLE Dec 82 Does Anybody Here Remember PURPLE I Puzzle Aug 74 Puzzle Stinky Pinky Aug 74 Puzzle Telephone Directory Oct 74 An October overlap ' Oct 74 Puzzle Telephone Recall Nov 74 Answer to An October Overla J E Dec 74 Puzzle Citizens of the World Dec 74 Puzzle Secret Messages 'Military Cryptanalytics' Jan 75 Puzzle Crossed COdewords '__ -r 1 Feb 75 Letter Citizens of World puzzled ' G P Feb 75 Puzzle Can YguMakeOutl he Name c J Apr 75 Answer to Can You MaK e Out the Name May 75 Puzzle How Many Words in 'CRYPTOLOG' Jun 75 Puzz1eJ c Aug 75 Puzzle CRY-PTO-LOGrolling Oct 75 NSA-Crostic No 1 A J S Jan 76 NSA-Crostic No 2 A J S Apr 76 NSA-Crostic No 3 A J S Jun 76 NSA-Crostic No 4 A J S Oct 76 NSA-Crostic No 5 A J S Nov 76 TEXTA 'Word Seek' I Jan 77 NSA-Crostic No 6 A J S Apr 77 NSA-Crostic No 7 Chauvenet L R Jul 77 NSA-Crostic No 8 A J S Jul 77 Puzzle Match Them Up ------ Sep 77 NSA-Crostic No 9 Williams D H Nov 77 How Many Afr ican Countries CariYou spot 1 1 Nov 77 NSA-Crostic No 10 A J S Jan 78 NSA-Crostic No 11 Williams D H Feb 78 Puzzle Three Holes Mar 78 Answer to Three Holes Mar 78 NSA-Crostic No 12 A J S Mar 78 Word Seek ' Apr 78 conoces Bien la Geografla t Apr 78 NSA-Crostic No 13 WilliamsD ij Apr 78 Prizewinners in Three Holes May 78 NSA-Crostic No 14 'Sardonyx' Jun 78 NSA-Crostic No 15 A J S Jul 78 NSA-Crostic No 16 Williams D H Sep 78 NSA-Crostic No 17 r-A J S Oct 78 Know Your Geography I Oct 78 Letter NSA-Crostic 1 May 83 OCt 78 Letter NSA- crost ici Filby V R Oct 78 NSA-Crostic No 18 Williams D H Oct 78 puzzle Who and Whom L- --------- Nov 78 NSA- Crostic No 19 Williams D H Dec 78 NSA-Crostic No 20 Williams D H Ja n79 NSA-Crostic No 21 Williams D H Feb 79 Know Your Geography Feb 79 NSA-Crostic No 22 D H W Mar 79 NSA-Crostic No 23 Salemme A J Apr 79 NSA-Crostic No 24 D H W May 79 NSA-Crostic No 25 D H W Jun 79 NSA-Crostic No 26 D H W Jul 79 NSA-Crostic No 27 D H W Aug 79 NSA-Crostic No 28 D H W Oct 79 NSA-Crostic No 29 D H W Jan 80 NSA-Crostic No 30 D H W Apr 80 NSA-Crostic No 31 D H W Jan 81 NSA-Crostic No 32 D H W Apr 81 NSA-Crostic No 33 D H W Oct 81 NSA-Crostic No 35 D H W Nov 81 Cryptic crossword - - DeC 8L SA crostic No 36 D H W Jan 82 SA crostj cN 37 D H W Feb 82 NSA-CrosticNe 38 D _ _oH _ _oW __o r Mar 82 Cryptic crossword I ' Mar 82 Letter Strangest Bust of the Mont ' DibbenA P L Apr 82 l' Ac-CI tic No 39 n H W May 82 Bust Answer 1 -- - --- - 1 May 82 NSA-Crostic No 40 D H W Jun 82 NSA-Crostic No 41 D H W Aug 82 NSA-Crostic No 42 D H W Oct 82 NSA Crostic No 43 O H W E O 1 4 c Nov 82 NSA-Crostic No 44 D H W B L 86-36 Dec 82 NSA-Crostic No 45 ' 1 os P L Qualifiers Nov 76 Clarity Thy Name is Qualifier Mollick J J EO 1 4 c P L 86 36 Radio Electronic Jun 791 I M R 1 P L RAGPIE Jan 78 But Why Do We Do It 1 RAP I DTRAN 75 RAPIDTr--JI - UI -' T- ---' Readiness Jul 79 Naval Readiness A Basis for Comparison 1 I Reference Aids Sep 77 JPRS Language Reference Aids Spiegelthal E S Reflections Nov 74 Reflections m a Tr ans lator 's conf rence ' __-- I CRYPTOLOG S S8RS'iP 86-36 ---------------------- ------ - Rad i O F ng e r rintin Apr 80 AIT 86-36 Page 26 P L fMimfJfiJ 0 86-36 fA serlIff' ' SU1 dftf BftB etf lJPY 86-36 4009861 EO 1 gSElflSf P L 86-36 Jan 19 Reflections and V R Recommendations Filby Reminiscences Mar 77 A View of the Central Computer Complex in the Late 1970s and Early 19805 Phillips C J Apr 81 The Poet's Corner Aug 82 I Remember I I Nov 82 An Old Timer Is One Who rl---- Reporters Aug 74 Calling All S RAs -- SRi Symposium I Reporting Aug 74 A Spot by Any Other Name Filby VS OCt 74 Letter c JArtlcle LI - -_ Apr 75 Oral Reporting ANew Challenge for Apr 80 Help Wanted Engle T L Apr 81 Grading The Russian PQE I _ Jan 82 Letter Delaney Article Chauvenet L R Satellites ' '---------------J nr----J c W L J an78AnEarly NSA Proposal for Satell ite Remoting j j N6t 1 eW M ' ip L I W E ra79 tet rl I I Apr 78 Accentuate the Negative I Reutilization Oct 79 Another source I Russian Sep 74 Language in the News I oct 74 The I IExercise A Case Stu y in Special Research Analysis Filby V R Nov 74 Purity of the Russian Language-Slavophiles vs westernizers 1 I Feb 75 Review of 'Guide to Russian Technical Tr ansla t ion J by Salemme j I Mar 76 To Pull a 'ponYal' 1 I May 76 COMINT in the Russian Navy WWII I I o I I 8 6- 3 6 ing ' jLettertAug 79 SCREENWRITER Mar 78 A Linguist Lpoks at the Tube I R S Apr 78 Letter 11 - __ JIArticle 'Fennwatcher' _ Jun 75 Letter Volenick Article Dec 711 Jan 18 Letter j I Arr t ic l e L r--_ _ l Sep 78 A Matter of Style SDa Jan 19 Reflections and Recommendations 'Fiiby Dec 8 1 SleepWeU Your SDO Is On Duty V R Sawye r E t J Nov 81 Futuristic Reporting I Secure communications Jun 75 The Navajo CbdeTal k ers Resources Jun 79 The BaJ tic Encoqersf Reiskis A Sep 77 Knowledge ResourceM c Dag ement at NSAl I 86-36 s e nie t n adi f I---- I NSA I P L Security Apr 7 The 'Ice Age' and International Secur i tY l j Jul 77 Let ter Westwood Article I__ Jl l179 Source protection Our Agency I s Insurance policy 1 Dec 81 Sleep Well Your SDO Is On Duty Sawyer E L Jl l' P L I Self-Paced Instruction Aug 4 Self-Paced Instruction The Future is Now I I on Apr 77 Some Thoughts the Russian PQE I I May 77 Why Are These People Smiling Jun 77 Letter Clark Artisle1 J Jun 77 Old Russian Manuscript Ciphers OL4 tc L 86-36 L---- Ic J Sep77I D L Nov 77 Backing into Language I Acquisition Dec 77 Letter L_ rA iC le j Jul 78 I Remember SPELLMAN Salemme A J Feb 79 In Defense of The Indefensible Notes on the Russian PQE Tetrault E W Jul 79 Gears of the Mouth 1 I Jul 79 Russian Handbook of Spoken Usage Vol 3 May 83 SIGINT1990 Sep 82 5 IGINT 199 lJ Part o n ell Oct 82 SIGINT 19 90 Part Two Nov 82 SIGINT 1990 Part Thr e 1 J CRYPTOLOG Page 27 I P L 86-36 1IA11BJ 8 gIl SSIUJ1'i' snMBll315S eJ115JPY 86-3 EO 1 4 c ero 4009861 Sf leM L 86-36 EO 1 4 c P L 86-36 P L SIGINT 80s Jan 81 SIGINT In The 80s I I Lutwini ak W TwoV i Ws 76 A Soviet View of NSAd 76 Russian SIGINT and Electronic -- --- - I uotasfor 1ec ---------- '1 SIGINT Alert 78 F j P L I sper Reporting I EO 1 4 c 7880 Sov iet L sCOMI_N_T_ _a_n_d_th_e_C i v it Wan 86-36 P L 8 6- 3 6 P L 86-3 I EO 1 4 c Space Program Jan 76 63 Days The sovietJ ik sp 8e ijo c _ _ J Dec 76 Apollo-Soyuz 'l'esfPrOjeet 1 J E --- 1 s rn r 1 -___ Jun 77 More about More about the NSA SIGINT Summary Boucher M J Signals Processing Oct 74 The Mission o the Sjgnals rOCeSSin9 Requirements Panel Language SIMP Jun 75 Golden OlQie SIMP Tables Software Nov 75 DESKPAD A Programmers's Tool 1 H 1 - _ May 78 The Joys of UNIX D'Imperio M E Nov 78 Formatting PL 1 Source Code 1 '1---- K J Feb 82 Software - AAAS 1982 --- ----_ I May 82 Letter PJa intextd f O W D O rY O U S - pe l l p__e_k_in_g_ _ - ' ---JF I Jun 78 Jul 78 SeP 78 OCt 78 C H Feb 79 Jul 79 P S I P L P S 1 f r----l P L 86-36 Shootout at the SIGINT Corral l-----J' SRA Aug 1 Oct OCt in 83p J5 RYffi 9L9l 8'BeRl3 P P L 74 Calling AllSRAsT-- SRA Symposium 1 I I 74 Letter Sapp Ar icle 74 Ther JExercise A Case Study Special Research Analysis Filby V R 86-36 May 86-3 EO 1 4 c Never Again Gurin J P L 86-36 I Remember SPELLMAN Salemme A J Letter Salemme Article Williams D H o More B S BeJore Spellman I Spread Spectrum Jun 82 Amateur Spread sPectrum I W L Dec 75 The Great Soviet Shipbuilding Mystery Williams D H Jan 76 63 Days--The soviets in space P L _ SPELLMAN Jan 79 SOLIS A Vehicle in Search of an Engine 1 May 79 Chapenko Shapenko What Difference Does It Make I 1 SOLITS Nov 75 In Praise of SOLITS Grant L G EOL4 c EO 1 4 c P L 86-36 Spanish Aug 74 The Language of Beisbol in Everyday Talk i 1 lSantiil9o-0rtiz R A L Sep 74 Language in the Ne s Jafl75 Tile Case for COMINTReaders I_ _ _- H G Aug 77 Telling I t LikeltIs Salitiago- Ortiz R A 86-36 Special Researc h DeC 7G Why Can't T eyDes ign a Good SR Test Bjorklund K EO 1 4 c P L 86 -36 Spellipg Jun 77 vich Iss R-r-right A J S Dec 78 How Do You Spell peking r-I--- - Dec 781 SOLIS EO 1 4 c P L 86-36 I J U l_7 8_T h e s o v ie t feneral St ff I se78 _ II I SIGINT Summary Mar 77 More about theNSA SIGINT Summary Signs Jun 79 Sign Soviet Translators Feb 78 Soviet Navy ColTlman LPost Sy teJ'nf P L I SIGINT Handbook Jan 75 The SIGINT Users' Handbook or What's an ISHTAR I 1 I 1 Warfare 86-36 SIGINT Exploitation Apr 78 SIGINT jXPlopatiOn 199011 --_ _ Jul 78 Letter Articlei Gurin J SIGINT Reporting Aug 76 Initiatives in SIGINT 86-36 Page 28 EO 1 4 c P L 86-36 86-36 4009861 P L Sep 77 Whither the SRA 11- ---' Telemetry Jun 76 Computers In The ELINT and Telemetry Businet s s I I SRI Apr 77 Flash 115th SRI Located Salemme A J 01 4 c L 86-36 reb 781 I ' 86-36 ----Jn ---2 iL _v_7_7_' P L 'IP 86-36 Telephone APr7ErTeiephone roblemHere L Mar 82 A Brief Treatise on Five Laws of Telephonic Communications Nolte W M Swahili Sep 74 Language in the News 1 __ Jan 7 SYMBIOSIS Jun 7S Project SYMBIOSIS 'Anon ' System Design Jun 77 Human Factors and Systems Design An Estranged Relationship I I System Development Mar 82 Towards Better System Development 5 UNNA I E6i JL - ' c ' -_ Terminology P L 86- 3 6 Aug 74 A Spot by Any Other Name Filby V R Nov 74 Data and Definitl 'ns cal i ng Thl ngs by Their Rightful Names _ 1 Aug 79 An August Baudy L Nov 8 L$ay What You Mean Gaddy D W J A ea FieldNafi 9 di 9 Conventions iiii p L 86-36 T 4TDB The'l'E T A tJa aBase j ksonW J T-Vision Jan 79 T-Vision The Reference Analyst's Medium of the Future I 1 Tale Sep 74 Golden Oldie King Eusyb and Queen Deodi l Oct 81 The Stairwell ociet y I ' - - 'I I Jan 82 Letter Stairwell Society Artlcle 1 Telex Nov 76 T TA 'Word Seek' I lulTS YeGads Another ount1 y Trigraph System Oct 78 Letter I rticle Pattie M T Dec 81 TEXTA Wh - -a ' t I -s I - t Where Is It Goirlg I Sep 77 The c Jseminar Program 1 1 W L Target Location Jun 78 Golden Oldie unidentified Unit at Unknown Location Third Party Apr 81 Third Party Relationships 1 __ TDY Jan 82 All The Alligators Aren't On Sport Shirts Teaching Methodolgies Jul 77 The Transcription Skill Concepts and Teaching Methodologies TIDE Dec 81 TIDE A Brief Time Usage May 82 Human Factors Corner Data Gathering How Do We Spend Our Day 11- ---- 1 Technical Exercises Sep 77 'Simonoff Says ' A J S Technical Information Jun 79 The Story of MOSES 1 L -_ _---il TIPS Sep 76 TIPS is Still Alive and Well ---' I Technical Manual Nov 74 Guidesmanship--or How to Write Technical Manuals Without Actually Giving Anything Away Mar 77 Revised Technical SIGImManual in Preparation Filby V R Traffic Analysis Aug 74 The New Traffic Analysis Glo$sapy Sep 14 Gary's Colors Garofalo C A Dec 74 A Approach to Callsign Analysis Apr 75 A CommCh lnge at Ramasun Station Technical Support Oct 81 Technical Support K J CatalOgS ' - -_ _ J I I May 15 'l'p Hand idl n f CAl Mason F O Jun 7 iMoreoIlSquaringthe Page A Crypto-TA FunctionbM ls F Aug 7 5 Abdul and His 40 May 83 CRYPTOLOG '1'al1 p Mason F O P L 86-36 Page 29 I1fdfBI B VIA SenIll'i' SHl tltIB 8 ell Y EO 1 4 c P L 86-36 SHEURMl'l' Auq 75 processin Apr 81 Traffic Analysis Specialty Without Portfolio ml mmmmmmmmmmummm ' H il' P L communications Lr r------------- I Oct 75 Automation of a Til Process Murphy T Nov 76 How to Make a RailroCicLPisappear f I Jan 77 Letter Mason Article I Apr 77 Letter Mason Article Boucher M J Apt 77 Letter Mason Article I Apr 77 Letter Mason Article McGrillies J R May 77 A Story with a Moral Jun 77 Letter Mason Article Oct 71 Which Numbering System Should We Use 'Asken' Nov 7711 Little Til problem - Dec 77 I - - 'J J Feb 78 Collection-Supp6rt Til is Not for Everyone 1 I Apr 78 Telephone Problem Here I Ma 78 TA 1m lications of FCC Proposal I __ _ _ Nov 78 A Small Problem __ Jan 79 Second Sighting 'Donym' Mar 79 Let's Not Lose Our Til SkiIIS I 86-36 Training AlIg74SeU 'Paced Ifistruc tioru'l'heFufure is NO I 1 Nov 74 llewTrends inthE 'l'eachingof Cryptanalysis IL - -- __ I Apr 76 How Do Adults Learn Language I I May 76 Hypnosis and self-Hypnosis in Language Learning I I NOV 77 Backing into Language Acquisition I I Dec 77 What Made Them Do It Language SelfStudy 1 I Dec 78 AgencysuffilUermLall91 1ageStudy I I Eh ooooooooooooooooooooo ' L P L- 0 18 6 _ 3 6 Jun 7g Teaching Computer ScienceTb LinguistsdL EO 1 4 c Oct 79 NCS Summer Language Program P L 86-36 Trat scr iber ' nalyst May 79 Transcriber Analyst Relations Miller D E P L 86- 3 6 u - G I Apr 79 A Somewhat Larger problem W E Jun 79 Letter I I Article B orklund K J u n 7 9_T r a f f 1 'C Analysis of the Future I Transcription Mar 76 To Pull a 'Ponyal' 1 Apr 76 Computer-Aided Transcription of I I I uuuuuu May 77 Why Are These People Smiling EO 1 4 c I I P L 86-36 Jul 77 Which Tape Has the Intelligence Project I I llurin J Aug 77 Telling It Like It Is Santiago-Ortiz R A Jul 78 I Remember SPELLMAN Salemme A J Qct78 u Can't Tell the Wh at from the Cha ff wJthout a Program Gurin J MCiy79 Run This Throu h Your Transcription Machine May 79 Whe ie ' 0 G O o T rii rlsc t it e r s CQme From Immmmmmmmmm I OCt 79'l'11 ZOOO--Year-Old'fI an e t 'ipel ' P L 86- 3 6 Jul 79 Letter Bjorklund Letter Jun 79 Broglie E F Jul 79 Letter Bjorklund Letter un 79 Buckley D I Jl l179l Aug 79 Letter Buckley I I I 0 I EO 1 4 c Mar 76 A Comparison of NSA and ATA L 86- 3 6 Certification Standards 1 J Jun 76 Notes on Translatip n f r o m t h-e-C hinese Traffic Analysts Aug 76 Integrated Analysts for Asia L W D Jan 77 Letter 'Firebrand' Letter on Article __ ---- Apr 77 Letter ArtiCle D Apr 77 Letter Article Jul 77 Letter Article Jul 77 Letter Article r Oct 77 Golden Oldie Analyzation of Data Curtin R Mar 79 Attention Military Traffic Analysts rer 80 A TraffiC Analyst Looks at Computers I I I I Sep 76 Semantic Voids DOn't Shoot the Translator Mason D Nov 76 How to Make a Rail oad Disappear 1 I Apr 71 'It's Got 1 0 Get Out Today ' Blank F Nov 77 WhatEver Does 'However' Mean I pr 78 ATA Letter l Q P esident Carter 'J 'inpley L MaY7aU 1 c le -a an Wantsa YOU I May 83 CRYPTOLOG SBeRBIf Page 30 P L 86-36 yn tTP r i HI n OJlltl QIWJUilJ5lil eu JPY 4P L Jul 78 Has It E veJ Been Translated Before 1 ' ' Nov 78 Jul 1 9 aul 79 jan 81 All 9 82 Letter I IArticie I -Gears oft e Mouth asleYD Seminar on Transiation probiams Translator In Your Pocken Linguist Machine f I ---J1 J Dec 75 Mum's Still the Word 'The Ut Secret' d I ' Dec 75 weapon That Helped Defeat Nazisi Winterbotham's 'The ULTRA Secre t i t r tbY P W Feb 78 More Beans Filby V R iii o Oct 78 Feeding the Germans Misfnformatip i Book Review Filby P W Tr rislqtiori qrading Mar 7 6 An Obj ective Approach to Scoring TranSlat ions 1 o '1 Aug 7'6 NSA's System for Grading Translations I I Translatbt AUg 14 Nice Busman's Holiday for One NSA Elmploye Pudley B NOV 74 Reflections ona Translator's Confel ende J 1 Oct 76 work quotas for Soviet Translators I I UNIX How To Mar 82 But Life IS Supposed To Be Hard I May 82 Letter 1 Sep 82 More Free Goodies May 79 Chapenko S 1apenko What Difference Does It Make I_ _ 1 Trigraph Jul 78 Ye Gads Another country Trigraph System I L 1 J Jul 79 Letter Linguatriv a I Apr 80 Geographic rivia _ _ UNIX Shell Apr 82 Shell Game 1 --_ _-- - I oo oo o May 82 Letter Shell Game Article o o o Jun 82 Letter Shell Game Article L j J W Jun 82 Letter Shell Game Article P A _ Jun 82 Loooong Shelr-l nL Aug 82 Shell Game I Nov 82 Letter Shel l G am--e--Air t l C- l e il _---J Dec 82 Shell Game AJSQUEi I ' - -_ _ Dec 82 Shell Game Counter W E S ------ Dec 82 Shell Game PWB WHEN IL UNNA Jan 75 Uml I DRS I 0 Apr 82 Word Processing In A4 Sep 82 TSS Revolution 1 J '---EO 1 4 c 'l'ypewriter P L 86-36 Aug 75 Typewriter Random -- A New Look I rticLe 1 G 1'ss tJO J Transliteration o Jun 76 Transliter tion or Cyr Hlic 11 true Base May 82 True Base Two Tales I r Nov 76 GOldecnOldie The Things They Say Mil ler D o E oo Jul 78 Is A Translator a Professional 1 Jan 79 w W II apanese Translation at ArHngton Hall Station Trivia Jun P A 79 Letter 'Sixth L 1gUage' UNIX May 78 The Joys of l1NIX 1 i Nov 81 How to CreateAU et-Unfriendl yooo oo ooo System I _-- -_ I o o Dec 81 In Pur sui t of Fast er Hor lesU10L nger Women Older Whiskey and More Moneyfd D L Feb 82 Human Factors Co ner Some Advice to J Users of Unfriendly System I Feb 82 Letter UNIX ArHcle Oct 82 Human Factors Corner Text Editops 1 1 Oct 82 What's The Good pass Word I __ I Dec 81 The 1981 URSI XX General Assembly I I UTENSIL May 78 Project UTENSIL The DDO Data DictionarY Directory I Oct 75 Letter Typewriter Keyboard Validity Dec 75 What Are We AQout Fragments Figments or What 1 I Feb 76 How Do We Know 1t's True ilby V R I Letter Apr 76 Letter Comments o on What Are We About 1 I Apr 76 Letter What Are We About tticle T Organization Mar 78 Some Background on the CiT Merger Smith F Sep 78 T Establishes Human Resource Development Panel ULTRA Dec 75 A Personal Comment on Winterbotham's 'The ULTRA Secret' Tiltman J H May 83 I I I Apr 76 On Being Tru hfuI Gaddy May 76 Some Principles pf Cover CRYPTOLOG Page 31 D aHa P L 86-36 ero 4009861 EO 1 4 c P L 86-36 P L I 86-36 I Deception Nov 76 Clarity Thy Name s QU fier Mollick J J Videocasette Apr 81 DIA Videocasejrte Program Video Encryption Jan 82 Video Encryption A Report 81 q I Weasel Words Oct 74 Golden Oldie An of Weasel Words II Oct 75 NSA in Vietnam Proud and Bitter Memories LI __ I Oct 75 One Chance in Three--But It Worked I o Oct A 75 The Danang Processing c ter 11 - _ Oct 75 The Do Xa Pads wiley E Oct 75 1 F ' - - - - - - - - - - - - - E O 1 4 c Jan 76 Leo in October Murphy A I Mar 76 Letter Proud and Bitter Memories Article Mar 76 Letter Proud and Bitter MemQries c J Article I Apr 76 Jun 76 Mar 79 E L May 79 M L 86-36 Unoffici Glossary EASCON rom Vietnam Feb 75 The Gulf of Tonkin InCident L __ W D Jul 75 Re-psychling the Code Clerk LI Aug 75 Twenty Years of Transposition L J E Oct 75 1972-1973 A VietnamOdysseyEO 1 4 c Stepp L C LI_ _ _ _ I Oct 75 Computers Comms and Low-Grade Ciphers l -- L l l - waveguide Analysis May 76 Waveguide Analysis I ___I _---- One Day in Danang I - -- - JI Letter Cameron's Article Murphy A Pursuit of thel II Project HELIPAD An EPitaPh IL - Working Aids Jan H G 75 The Case for COMINT Readers lL- _ writing Nov 74 Guidesmanship--or How to Write Technical Manuals Without Actually Giving Anything Away ThaI W E - May 77 Plain English II - --__---- __ Aug 77 Let Me Repeat--And Make Myself Perfectly Clear Jenks P Jan 78 The Joys and Frustrations of PluralDropping A J S Dec 78 'No No Nanette ' Means Yes A J S Mar 79 Fairbanks on English Fairbanks S Apr 79 More Fairbanks on English Fairbanks S May 79 Moree Fairbanks 01'1 J'lgH l1 r fairbanks S P L May 79 More ThanWotds Can Say Gurin J Jun 79 How Are Your Stamina Fairbanks S Jan 82 The Literary Bends Murphy A I Feb 82 Native Scripting of Languages l I ' '''' P L M a r B 2 L e t t e-r- --T h e- L terary Bends Article 86-36 86-36 MurptJyA I - I Apr 78 Looking at Mr c J1'----_I N a s1s 1 a_l_O_g _ _e_ B _ _elween MS User and Dr EO 1 4 c P L 86-36 VORD Mar 79 VORD is a Better Idea IL- -- Voynich Manuscript Aug 75 The Voynich Manuscr ipt-Third Theory Miller D E Apr 76 The Voynich Manuscript Revisited WARC-79 May 78 Callsigns and ARC-79 L riolll fmaatstept03 L__ Subject SHELL Reference CRYPTOLOG March 83 page 32 _ Warnings Dec 76 But It LOoks Like the Real I Mar 77 Letter Bunker Interview 11 _ H L Dec 82 Development and Correlation of Indicators LI 1 WARSAW Pact Jul 75 The Warsaw Pact 11 - _ _ I P L May 83 EO 1 4 c P L 86-36 CRYPTOLOG 86-36 I T441 963-5553ro 1 4 c P L 86-36 Page 32 UMIBl JB VIA eSHHl'f efl ti ffib elNty aCID 4009861 QUPIBl lff ' Ar E01 4 c P 1 86-36 OUT OF MY PTH IUj May 83 CRYPTOLOG Page 33 4009861 SB8RB'i' The Intelligence Watch Officer U L EO 1 4 c P L 86-36 W I ant an intE r e s t ing job too Do you to be busy --sometimes busywant to eat even when you've brown-bagged If you are interested in cur rent U events and want to know what is going on in the world or what is going to happen next you may be intet sted in the Intelligence Watch Officer IWO position in the National SIGINT Operations Center Ii U The author served as Intelligence Watch Officer on Team 3 P33 in NSOC from September 1981 to January 1983 1 May 83 CRYPTOLOG Page 34 8BeftB'i' 86-3 4009861 SS8R13'f U If yo are interested in more information on the lntelligence Watch Officer position in NSOC contact'r5 x3265s EO 1 4 c P L U During non-regular duty hours the IWO acts for T5 in receiving and responding to requests for information from NSOC field sites and 24-hour NSA work areas The IWO attempts to answer all requests instead of referring them to a day shop for action This frequently necessitates trips to the NSA Geography and Map Library the Main Library or to Central Research to research and retr ieve a map or citation to answer the query The IWO has a one-way pager to carry when away from NSOC for any length of time--he maintains constant contact with NSOC when necessary to receive new requests or operational requirements 86 36 Sir I would like to raise a full glass in a toast to n his article concern- P L EO 1 4 c P L 86-36 May 83 86-36 CRYPTOLOG Page 35 I lfe JUt sellB'' ' ElIWR'BI S eUI Y ero 4009861 L he traffic analyst finds himself turning to data systems because he often has mountains of data to examine because the people who receive U TA results usually want their information very rapidly and because almost all the data the traffic analyst wants to see is already inside a computer somewhere T U Most traffic analysts who try to look at data systems develop a kind of schizophrenia On the one hand the TA data that comes in today has to be processed and analyzed today b cause there will be another batch of data coming in tomorrow This means that the traffic analyst has to use today's data system to handle today' s data On the other hand it does seem to us traffic analysts that data systems people would much rather talk about tomorrow's system--the one that isn't here yet the one that won't have all these glitches and problems that today's system has U The traffic analyst who is in the trenches on a current operational problem would easily trade all the glowing promises of some brighter tomorrow for a quick fix on some of the glitches in today's system that will keep him from bleeding to death right now That isn't my subject today--I really want to talk about the future But as I thought about standing up here in front of all you data systems people I couldn't resist putting in a plug for the working traffic analyst he needs your help both today and tomorrow May 83 This paper was presented at the November 1982 Meeting of CISI THE PROLOGUE IS PAST U I might be well to begin with a little history or at least history as I remember it My first recollection of what we now call data systems was a' lot of 80-column cards and a card sorter That was about 35 years ago Watching those cards go through that sorter was rather hypnotic The possibilities seemed limitless then--if we could only find a cheap and easy way to get the data onto the cards' I think the equipment was called Electronic Accounting Machines EAM and the people who supported the traffic analysts were called Methods Analysts in the 1940s and early 50s U Since our data cons isted of a matr ix with 80 columns and many rows one row for each card our output consisted of that same matrix with its columns and rows transposed in some way Later we added the ability to look up words or strings in a dictionary and insert the result back into the matrix CRYPTOLOG 8rJeRB'I' 8P81ll3 page 36 86-3 SElSRB'f SPBHEI 4009861 U Many years and computer systems later in the mid-1960s this was still the primary data systems support to traffic analysts a transposed matrix now often wider than 80 columns with a dictionary lookup There were attempts to go beyond this Most of the things we tried were made to fit one specific problem and never developed into general TA tools We developed ponderous monolithic record formats whose structure provided a special place for each variety of data we thought we would find in the traffic What I remember most vividly are long soporific meetings where all we ever seemed to talk about was what format the data was going to be in We spent untold amounts of energy and resources getting all of our data into these unyielding user-murky systems and there was often little energy and resources left over to develop any user-friendly output U It is still possible even today to see analysts sitting down with computer output and handlogging data from that computer output onto a form for their own personal use In at least two areas one might then see that same handwritten log being used a little later to punch cards for further computer processing WHAT IS THE TRAFFIC ANALYST TRYING TO DO U The picture of ally wants looks when knows what then he is and report traffic analyst is trying to draw a his communications target He usuthis picture to show how his target it is operating normally Once he his target 's normal behavior is in a position to detect variations them to intelligence consumers CONTINUITY U Traffic analysts are usually looking for something they call continuity When faced with a target that has daily-changing callsigns the traffic analyst seeks to learn which of today's callsigns matches what callsign used yesterday ALONG THIN MATRIX The form in which the output was delivered to the analyst was often decreed by someone remote from the analyst--someone who never had to actually live with the output--and it was rarely if ever changed to fit the current needs of the local problem or individual analyst May 83 DATE -L 'SO'tt 1 ST' l A6C bE F G-1I1 _s_ - - If I can say that the station that used callsign ABC on the first day is the same station that used callsign DEF on the second then I can say that DEF on the 2nd day is continuity of ABC on the 1st On the third day if I can say that GHI was used by that same station then I can add GHI on day 3 as another link in a growing chain of continuity Many of our TA targets do change their callsigns frequencies addresses and other features on a regular basis They do it to CRYPTOLOG Page 37 SSSia3'i' spanH aCID 4009861 make collection and identification more difficult and it is the job of the traffic analyst to defeat these changes by the development of continuity TWO KINDS OF TRAFFIC ANALYSIS U There are two forms of traffic analysis on most problems development and maintenance To borrow an example from cryptanalysis the attack against a cipher system often goes through two phases J first diagnosing and recovering the general cipher system and of second exploiting and processing the recovered system which often involves solving daily keys or settings So too in traffic analysis one can consider that there is a development or recovery phase and a maintenance or exploitation phase which mayor may not include product reporting However in traffic analysis the two phases often occur at the same time U The more territory one conquers the thinner the center of the oil blot becomes The more communications structures one recovers the more continuities there are that now must be kept track of As the maintenance effort grows it will use more of the available resources draining them away from the recovery part of the effort and at some point it will have absorbed enough of the resources so that a point of no growth is reached and for all practical purposes recovery of new structures stops If expansion doesn't stop the center of the oil blot will break if development doesn't stop the maintenance effort will fall behind and begin to lose track of continuities which will then have to be discovered and developed all over again This tension over resources between maintenance and development is similar to the one between software maintenance and software deve lopmen t 100% Maintenance J W U In some ways the traffic analysis process resembles a spreading oil blot Out on the edges new target territory is being conque red new target communications structures are being discovered and cataloged new methods of identifying and distinguishing various communications are being developed But back in the central part of the oil blot the territories previously conquered must be kept track of the continuity of target communications structures previously recovered must be maintained o 0 J o J w a Development 0% TIME - OIL BLOT Recovery Maintenance outer edges center May 83 U Sometimes the personality of the manager plays a part in just where this point of no growth takes place Some managers are more at horne in the settled stable atmosphere of the center where things don't change much from day to day These managers tend to concentrate their attention on building a smooth-running system at the center and may put a larger proportion of their resources into that area so that the no growth point is reached more quickly Other managers thrive in the rough and tumble frontier atmosphere out on the edges of the problem where each day is likely to br ing some new and different challenge These managers tend to concentrate their attention on the recovery effort sometimes at the expense of the more humdrum maintenance CRYPTOLOG Page 38 aCID 4009861 TRAFFIC ANALYSIS GOALS U From the standpoint of the two kinds of traffic analysis--development and maintenance --we can express the general goals in the following ways 100% Maintenance i TA DEVELOPMENT GOALS en w ex J oen Reporting w a Development 0% ' - - - - - - - - - - - - TIME - REPORTING U Some traffic analysis problems have a lot of potential for reporting--for providing the intelligence consumer with a blow by blow account of what the target is doing Targets that involve ships and aircraft often have this potential because they move around from place to place and the analysts often find much of their time taken up with reporting which ships and aircraft were active today in what areas and performing what missions Where this reporting potential is high it tends to draw off resources from both development and maintenance Managers whose problems have a strong reporting emphasis especially time-sensitive reporting will generally try to pull resources from development rather than from maintenance because losing the continuities means losing the raw material for the reporting effort Losing the development effort is generally seen as the lesser of two evils U As an aside I should say here that the reporting side of traffic analysis is generally well ahead of the technical side in the use of computers Since my primary interest in this paper is the working-level traffic analyst I will be concentrating on the technical side and I do not propose to discuss the reporting aspects of TA except as they touch on the technical side May 83 U We rarely collect or analyze all of the communications of any given target We are almost always working on a sample of the target At any given time there is some residue of the target that we do not maintain continuity on and bits and pieces of that residue find their way into our unidentifieo or search pile--the file of incoming traffic which looks as if it belongs to our target but doesn't exactly fit any of our known continuities Development TA concentrates on that pile trying to dig out new target nets and continuities This unidentified pile is almost like background noise it is always there whether we talk about it or not If we are still growing if the oil blot is still expanding then our development goal is to dig mOre of the target out of the unidentified pile If we have reached the no growth point then our development goal is to be able to recognize and develop any new communications that the target might put on the air-communications that ought to stand out against the normal noise in the unidentified pile TA MAINTENANCE GOALS U During the maintenance phase we want to be able to hang on to the continuity that we have already recovered We want to do this to support whatever analysis efforts are currently engaged on the target such as cryptanalysis language reporting etc and to support whatever collection effort is working against the target U To do the first support requirement properly we need to be able to correctly distinguish and identify each of our continuities as the traffic arrives at the point of analysis Le after it has been collected CRYPTOLOG page 39 P L 86-36 ero 4009861 EO 1 4 c F L 86-36 SEelHl'f SP6 E P L 86-36 will always be some mixture of thes two forms of traffic analysis We also need to consider that a TA problem can quickly change from one form to the other HOW CAN COMPUTER POWER BE APPLIED TO THE TA PROBLEM S ee6 In order to consider how the power of the modern computer might be applied to traffic analysis we need to look at the model of TA that emerges from these two phases development TA and maintenance TA Although I have described them as if they were distinct and separate they really ought to be thought of as a conjugate pair because they tend to occur together on most problems It is also possible for certain problems to be best described as a hybrid of these two forms during the war in Vietnam one out of every three pieces of intercepted traffic was unidentified largely because of the rapidly-changing nature of the target The point I want to leave with you today is that any attempt to provide the traffic analyst either here in this building or anywhere in the world with a Traffic Analysis Workbench System must reckon with the fact that the problem he is working Problem type We decided to see if we couldn't find a way for computers to help us with the more stable maintenance problem I remember spending several weeks laying out the logic and processes on the problem And I remember being told at the end of the project that there wasn't nearly enough memory available to do what I needed MAINTENANCE DEVELOPMENT continuity-keeping bookkeeping Anything changed continuity-seeking pattern searching What's new slowly changing high -solved rapidly changing low -unsolved How dynamic Foreknowledge State of solution Control Interaction Techniques semi-automatic human-efficient target specific knowledge-based Worst case below the salt hands-on human-intensive human specific mix-n-match start from scratch A COMPARISON OF TWO FORMS OF TA May 83 CRYPTOLOG Page SBSRET SPBze 40 SilElRB'i' 4009861 A COMPARISON OF TWO fORMS OF TA U Let's look at these two forms of TAa little more closely How do they compare when we look at them from the viewpoint of providing today's and tomorrow's traffic analyst with a computer support toolkit while using a terminal workstation in an NSA worldwide networking environment U In development TA the garrison communications in our example we have a bookkeeping problem where the emphasis is clearly on keeping track of a lot of known continuities where we expect the changes in the target characteristics to be relatively modest J spee U In maintenance TA we work largely with what the target gives us A package of techniques to grapple with a callsign system may work well enough on a problem where the callsigns are the key to our keeping track of continuity but may be almost useless on another problem where the callsign system isn't solved and we must rely on other things such as serial numbers or addresses U In development TA on the other hand a particular technique may pull one new structure out of the search pile and then never again find anything The development analyst may need to continually devise new attacks and new methods to him the search pile is a featureless mass and it is his job to sort out the various pieces and find ways to distinguish one piece from another with some reliability U The maintenance TA problem probably needs a package that will where the technical means of keeping up with the target e callsign and frequency systems address tables etc are largely solved or understood and where we have good prospects of being able to project the appearance and behavior of the target from day to day U In development TA the training communications in our example we have a patternsearching problem where the emphasis is on sifting through masses of low-yield ore looking for something that forms a continuity J where the next success may look nothing like the last one and where the chances of finding that needle in the haystack may depend as much on the personality of the searcher as on the content of the haystack If we can't keep continuities Le are not able to then the target stays in the development phase no matter how much we know about it Someone once said that TA continuities take either 95% of our resources or 5% That number may not be right but the idea is Being able to keep track of the continuities is the key to whether the problem is development or maintenance in nature A dailychanging callsign system looks to us as if it is rapidly changing if we haven't solved the system but once the system is solved we then perceive it to be slowly changing It is a matter of viewpoint May 83 look over the incoming the day material for make reasonable guesses about tinuities including garbles flash a warning light at the traffic analyst when things look very wrong or when it is confused by something and provide a clean and readable summary of its results to the analyst for review con- It ought to keep up with both short-term and long-term trends and should be especially attentive about missing persons portions of the target which haven't been seen for a while U The development TA problem on the other hand needs a toolkit that will provide the analyst with a range of diagnostic computational and pattern-searching techniques that can be brought to bear on the problem in whatever mix the analyst needs at the moment fiORST CASE U I have shown what might be called the worst case for each of these forms of traffic analysis U In maintenance TA one sometimes finds that a problem must somehow be worked but that it has no real resources and not enough clout to get any Now in the best of all worlds where everything is done right and for the right reasons such problems should not exist If a problem is worth working on at CRYPTOLOG 81 leMl'f epee Page 41 4009861 SBeftB'I' SP9KB all it is worth the resources needed to get the job done However in the real world those problems that are below the salt will always be working with whatever support they can beg borrow or scrounge Providing a general package for such problems would pay for itself a hundredfold in the first few years At the minimum package needs to be able to r ing an alarm bell when the target starts to disappear or becomes more active or changes in some other way t l eee In development TA the worst case might be the situation where nothing is known That is not as uncollllOOn as some people might think I I U There are aspects of traffic analysis which resemble word processing and a good screen editor seems to go a long way toward putting the analyst in contact with his traffic letting him rearrange it and touch up the rough edges and garbles the way he the owner wants them It lets him look at the data before he decides what processing to apply to it It also puts him in a good position to generate reports about his problem especially the technical reports with technical data embedded in narrative text Ul A good toolkit similar to UNIX and the PINSETTER extensions is invaluable in provid- I i rh m lf r YSf i t y large scale manipulation of his traffic _ - - _- - - -_ __----II When we pull together n analysis effort for a sudden war or brushfire the analysts are usually drawn from other problems around the building it would be nice i f they didn't have to add learning a new system to all the other problems they will face on the new target Therefore the toolkit for such situations must be quite general and all-purpose U Many of the practical results of PINSETTER results that found their way into daily applications on specific targets were not limited to traffic analysis It became a regular occur renee to hear people from other cryptologic disciplines tell us that much of the UNIX PINSETTER ackage for traffic analysts was what they needed too PROBLEMS TWO 1 4 c P L 86-36 THAT NEED SOLVING SYSTEMS OR ONE What I have been describing so far may sound like two different systems but what I am proposing is one system with two parts I have already said that these two phases or aspects of TA occur together and I should add that on more than one problem they are frequently done by the same people New continui ties are recovered by the development TA process and then handed over to the maintenance TA effort to be kept track of Information is often derived by the maintenance effort that will help the development effort What the traffic analyst needs is one system that has enough flexibility for him to move whichever way his TA problem takes him It would also be useful if the language we use is one that isn't going to change every few years because some equipment in the basement is being upgraded U PINSETTER U Several years ago we began to work on the concept of a Traffic Analysis Workbench System with the covername PINSETTER Some of what I have descr ibed here comes out of that experience PINSETTER has been described elsewhere so I will not spend time on it here However I will share with you some of my personal conclusions about PINSETTER especially those which seem to be pertinent to the future May 83 U Among the many problems that need to be solved I would like to mention two Both of these are areas that are critical to the future TA Workbench System ARCHIVES U Some of our continuities form chains that stretch back to the end of World War II One of the things that Data Systems people don't like to hear is that we need storage for data whose lifetime must be measured in years and perhaps decades Some years ago there came a time when all of our incoming data went solely into the computers in the basement It was the culmination of the dreams of a number of people to take the raw traffic away from the analystl I don't challenge that decision It is history But I must say that on many TA problems around the agency there are no good records on our known continuities from that date forward unless there were analysts still keeping some sort of hand records The philosophy on most computer hosts is that any records not accessed within some period usually a year or less are taken off the system U Even if the data is put onto tape the medium will deteriorate Once on tape the data is out of sight and out of mind The software that understands that data will sooner or later disappear or be improved Nevertheless the analysts on that problem are still responsible for that period of time and may still have to field questions about their CRYPTOLOG Page 42 4009861 targets for that time period So far we have dodged this bullet but sooner or later we will have to face the need for long term archives INFREQUENT USE OF PROCESSES U The second problem involves the question of software that is only infrequently used For example suppose that one of our larger targets has a major communications change every five or six years The effect of this change is so great that it interrupts intelligence reporting on that target until the new communications structures are understood and recovered Each time the change occurs an intensive effort is therefore mounted to recover our continuities in the shortest possible time U In the old days when the special effort was over everything was bundled up and packed away for retrieval when the next change came along But how do we handle this now that we have modern data systems support After five or six years how much of the software is still useful Chances are that the data base has been changed as well as the host on which it resides U Another example might be the diagnostic techniques to attack a particular kind of callsign system Once the system in question is solved how should we preserve the software so that it doesn't need to be reinvented the next time such a system is encountered Suppose we don't find a similar system for five or six or even ten years CONCLUSION U I don't offer either my observations or my experiences as criticisms but rather as areas of traffic analysis support which need to be solved I have tried to avoid mentioning specific hardware or software except as examples A man named Bob Biles taught me long ago that users should never tell computer people what equipment to use U Perhaps traffic analysis has lagged behind other cryptologic disciplines in making full use of modern data systems But that is changing thanks to the patience ingenuity and hard work of many of you here today I still keep a supply of pencils around and I still have a pencil sharpener on my desk--but I have noticed that I don't really use them very much any more May 83 I Dear Editor I I I U In sympathy with countless NSAers who I through the years have been antagonized bafI fled challenged demoralized etc by countI less forms of human and machine language not I to mention the devil's own creation governI mentese I propose that our new OPS Bulding I 2A be christened--at least informally-- The I Tower of Babel l I I I JOHN J MOLLICK B41 I I I I I I I Editor's Note In keeping with the tradition of naming the streets aUditoriums etc in the NSA complex after outstanding individuals we could always claim that the building had been named after the late lamented Mabel Babel 19 -1979 one of the Agency's foremost linguists who spoke fluent governmentese as her native tongue Her classic work A Governmentese-English EnglishGovernmentese Dictionary now out of print is still the classic work in the field I I I I I I I I I I I I I I I I I I I I I I I I I Mabel Babel f I I ------------------I P L 86-36 I I SOLUTION TO NSA-CROSTIC No 46 I I 1r------ I Memo from the Editor of CRYP' 'OI LOG to CRYPTOLOG's Puzzle Editor I I I I I I I Our Jan-Feb issue will be devoted to the CIS I Essay Contest It would be useful to have a puzzle that has data systems as a base if you can find a suitable text Also I have been contemplating running an April Fool issue You might be thinking about that ooo CRYPTOLOG SBeRl3'l' SP9KEl Page 43 aCID 4009861 F8R 8FFII3Ii'rI HS 8U15Y N S Norway 1889--1960 was an aeronautical engineer active during the pioneer days of British aviation He was also a prolific novelist Many of his best works are set in the country which became his post-WWII home Mr Norway is Word V in this puzzle May 83 CRYPTOLOG 'I Page 44 F8R 8FFII311rI HSE 81ll5'l P L 86-36 4009861 EOR OfEiCiAL SZ ONLr May 83 CRYPTOLOG 1 g 4586-36 FeR 8PprerAb ass 811bY Pl-AUG 83-53-35267 uocxn 4009861 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