OCR of the Document

View the Document >>

National Security Agency, Cryptolog Vol. 3, No. 6-7, “Computers in ELINT and Telemetry,” “Transliteration or Cyrillic?,” “Comments on AG-22/IATS,” “The Goodkin Collection System,” June-July 1976. Top Secret/Confidential/ Confidential/Secret.

TOP SICltlf Wul il'CIUi il1 ttl IBID OO CB v W ' L l l i l ls I LllBall J f' tJU li t ll WWll B Eij UCDl1CD J£Dllllg Df1 1•••••• COMPUTERS IN ELINT AND TELEMETRY • • •• r itANSLITERATION ot·cYRiLLic1 G ·T d COt-1'-tENTS ON AG-22 JATS • 11 ••• l 13 • • • 18 Wl llll ' ' '' 'IT T l 'ft 'llll'C lll'llll'l l'ff '-------- 11 'Tl •• •• 20 •••••••• • • 21 ' • ZS 26 28 _ _ _ _ _ _ _ _ _ _ ______________ _ _ _ _ _ 29 NP IIGIII · ' · THPS WICl'Nltz CO TOINS conmelftP tor 55 IQ 1 Non - Respon sive r r tt I ncn± 7·z mg rm I srn a eJ D I so tr m mH r c s• M •r W13 1122 tip O O Id 1 I PL 86-36 50 USC 3605 1 Declassified and Appro ved f o r Release by NSA on 03 - 19 - 2021 pursuant to E O 13526 MOR Case #109432 OF SDI _ I Non - Responsive - - TAR SS S• III I - ·-· -- ' lt e1tct _ - - I - PL 86-36 50 USC 3605 I C•U r ITHE PUTER5 Ell T fltiD f ElEr lETRV SUS ES Chief W In order £or you t o callbrote wut I h ave to ta ' I vut to uko h vuy cloaT that t a• not • c omputtn' expert bave nev r partldpat in the dcu lgn of aey ccmpu on f havo 4d little cpent lonal e•rionce 11lth tbe care ud teodlna of the11 and 1 havo never 1ffitten sottllare I un tiad the opportu • nltJ to h lp deslp Utd wor vith a nusiber of syatea thllt tase cou utt-r• in lt r 1 iaportant VAT• · and I •• ln a JIO•I Uon 1 1 W lo •fleet tho arket lot coaputen i n tho BUMI' and telemeu-y ' bUSlne- 1•11 be lll H ctla rr tNrkt n dlJ•• not r o rJ thcaucl ves of enu r ndar-s ThJ med to ba roal shq lc 111 en ataffs too la'l'lo lluo 1114 there and we hAd several t ll'el of El INTI sooo Teclut lc za 1 l LJNl' tol O OpontiomU LIN so eieetronlc lf4rfAro oir o E lectrmc Slfl'POl'tNllaaurn srut to soce eirer rhina rhat lsn•t CONftff' 1· actua lly cal explain au that bbt l voulil tal o too Iona 011d lt bn't rca11y very ta portini except tbt lt contr lbuto to aoDo of tho co tusl op The ELlNT I'D Pina to talk about tod t ' b aostl ' f od for tvo tltlng cs rl _ -1111 ----- _ 1 111 cn don't intmd to NY 1111 lJ s very flieJe two thin an tht'll Sob s lntencttvo profouild or ti ' be afrai d ttlat you won't •l t p t t b all t b•t 'O I tmow • you ofton • 'nils kind of liLINI' la conc omod vith tbo doc •ll yaur taU or book an lntr-od Uon to • t emiutlo 1 ol the copabi lhy of tho l'adar l 1 vo rnhtcd that today Jus t to l uro m n of 5YHCIII It ls bw 1t on a stpah-M lly b • or 1 tt e lut it 11e s S IISts ted to • that a u3sunnent •pot Utty Today it b usually bri of int tio t t o el l il' llJIIJ tclCJJOtry mi ght ctlled Technical l LJNr cl in order 1 • U nart with tbat and Uep But wha cfo vu ao wen rou look ot a ndar7 it bHh TMn l 'Jl phHc into tti do• Unf'ortunatoly the trannlu lon fro11 tlla ndu 1Cdpt lol of $ currtn tct of c ocpute t op_pllca• and u n onl1 that 1 tions and ml wlth wtlu I sec as real trends ln this 4rOA 1 le -adl10t-um w a nR' and' r ti To DOst poaplo ELIHT l'P ll J Nl S GIi thGse ° t hin1s that eo oo to oxplol t ho tt'811staisJiorut -----------·-----··········· EO 3 3b 3 PL 8 6-36 50 USC 3605 lOR CiiGR l a wee cu au aau • 111• • EO 3 3b 3 PL 86-36 50 USC 3605 - - ----······ -- ----- - Islsnals 1ro11 the 1ysie11 e - can iet ot oiher we aTe usu111r hotter olt Vl'll lc t Ma ahray loob simple altd in concept 1t ie 11 1e real U t h tt •way of s oo onr lfc thinl vo eati o it lnfoffl3tion ' us r ex w wlt fOU 'IIIIIC Our ai aln job ii to sot that iafOPati oo • de• codulatc it cmter or dbpl•Y it tnd euclne i in som •Y which will help the co lty to decide lllllat r a• goin1 on and what n -'lll'OrtaJIC Generally sneakln1 ther are fewer ar eas of confusion in noacmclaturo 11'TOUrtclin1 tC Jeutl' ' ttovcver to SON people U 15 IUSt ELlHT To othen it b COMltrr To IOJIOt oaly that n uff whicll IS related 'nl'Wll'll 'twnrl lil o to lncl c 'line• liko bco cons certain lnstl'Ulllntat1on n • nsponden lnterfer netcr 1ip11ls and othor thin • l nto WMt i • called FIS Fonlp Jn5tl'Ullentatlon SIJn• ls ' For our purposos horo today it la sofo to think of 1111 or that as tel emet ry bottcr ftOW But nar bor wo 11usi usOllble dab i hat ls I llcel ' to be c 01 lected 0 ICt S tcna period of Uao and rot 111 n confidonca fo that data When we d DI witb telocr ctry wv arc usWJlJr a Lltue i etttr off A leut sueone h lnteiid · - - - And t • te1eue17 ere to CONINI' to 1ot SltlfNTI now o 'to C011pUtcrr 1 t sap o'f c AppH9eUcn• I l fo' u o •• dapead lift • • computers ln au e or nrr and io1aouy htt l Uguc I C0Uectu111 • flltocu · •Ope te1Uon 6 'It 11 l 8 so me en - 1c1 of Ucatlons in O rr tt e l ntelUsontc f i eld but 11 11 avolcl ta11tin1 t out Dau Stor1111c and' Support • Whlle sut'h It is important to undcHtand tho mpltud appUcatlons aro iltnlficant tbey an Nit nGW o '- • U r - et · I ho i u 111 l l r e a ll z e t ha t-_- 1'auch d iffcrcmt hom any co rploit fil lna ' problea You Shollld not lnfor that ••• happy with llhal goes on i n thh al'Ni Soaetlaes •I thlnk our Oftty solution would be a big ttrel I vbll that tcr acbody wuld find a fundaftnt al y cU ffeNftt way 1 0 file tnfonatlon I hOpo of 'OU IN workin1 thal problffl r ha¥' no ldeas to coiatributo· Operatlons • • oUIU'l f I not SIGINI' •• are l ncltldcd in tho box abo110 bee auso C1n l ntercstlna coqUna blsts bore l'ffbaps I t ls UAlque in tho Jntollipttec business We'll tow 11 oo tbt later in this rosenuuon t•a loot ai • __ ' '___________________ operation Fl1 1 vorr aoper•llied collect ion eo iort •re u sed •ch This eontinuos to lie a vo-r' tq cni a11t 11011N e aon exunslvtl ' ln tho collection ptiuo of ou ' of infonaation You P1 bo in oni ted 1n nad· wrk tllan b rrtnUy the HO ln OQNINI' Pao• lna what Dr Tor-dolll Mel to $111 ' llbout t alceuy ple have Nm oblo I Cl A J l ell iUI daU-hllndlln ' not voey lona O O It's rdc d I n tile nil Juno-July 76 • CRYP'IOLOG • Pase 2 JQA iCfliT 111112 n 111 0012 r a 1mm a en EO 3 3b 3 l P liGRiT PL 86-36 50 USC 3605 l - - ···········-------------------- • syatccas fho senerali cd scho 1111 that o •ee 1 hoN could be almost anything but 11 11 cl i• that it reprosents an ackancll dl ELIST collect Jon ' $ 'Stem You'll not lh t Ill parts of th thine pcrl op c11rn tho opc rat-or are undH • l'l lter control • ·' I tblnk lt b iGFr• sslvc It hu Men opuradng tor o 'cr t bn o yun Tho l'llnctions of the t o r arc cssonti l PUt anotllc'f -ay you could not do this lnt•ll1 goiice coll«tlOft Job in ny O ber WlY• Ith that si ple Tho c oaputcr ali'la lt posslblo 1beT o uo other functlon Lly slallar 111t I both tuger- and mall t lKtlh ln opcr•• tlon and plaMlnt lt lCS at lhh iDO f'lt ta e odm COMPUTER s lsni4 For 4 look Jt II sood claar ic proc eulna systesi lot's iaovo to tho tollllN 'rf a-rea Again I have II very •i Jllllfitd ldea of wh it const1tutos a proc essln s1 tea •· Dlilr st any t111na wl 11 flt into t hc block CF le 2 PrtOCESSlNG SVSTEII Let•a site one cf£or the lariest that 1 lh bo it isrked ••c oq uter sync in Fla l ls mde u of mew of 1 2 By •1 dcflnltlcm vo aennuly an nottrorltna on• lino 11Uh 1 be d at a Sou sort of bUEfer l na or scon10 • • pre1eni l y tape •· ls used What COCICS out Of a ptotOSSOr isn 1 t yot a pal'ti larly vsoful pieto of nteJ Ua ee or oven SIGINT Pet ple atilt need LO b1t Jawolved Ono of our major lnv•s1aents in th toloaot • 1'f•ptoee sslna busineH la • thlna '°Ucd f MIHM t could allow 'OU a picture of it but 110u people wllo work wlth c Oli plrte1'S have II proh• len you•vo Hon ono JOU1 'IIO soon tbc l all I Mnl N u blailt • sqlo SJGK -V ccm- ·puter wlth tbrae disc uni t• and ala rapid• 4CCOII dcr lico• IIADI Tho l lnd of job Wt 1t oes ls p ly best described by a lool et the input slaul It works Jlo tlY 0 a slanal which 11111 C ll Jmut-Jaty 76 • CRYPfOLOC • Pago l liP iliRif HtxPlf m n z m 21r1 e s ur j• i i i EO 3 3b 3 PL 86-36 50 USC 3605 wr eM 1 ' 11--------------- ---------- you'l O OOd •c doing it Q£illt • liko your help tool A i 4 wo•cl daes tint Don't forge the voluaol Soven1 Ji ' tms of coca ar•bl• c bllitv £1115 in Jbt ss11I5fDltv INt not for such_ A i ilmo Syvt01t My 1nalytl c exa11ples are allo bued on the slqJc concept sPIIMl la fia -'• 'ltie •11na1yt1c systelt - be o special coimlnatlon of bardwaro an• sof'tlllU C or sltiply softnn tor So IO nice Beneral-purpoH a acMne ne slie vulH a tat horo fl'OD progrrmullle Kcvlott-hclkard clesk or band calculaton to CDC-7600s ANA LVflC SVSTEII Fi1 3 Let's look at uo oira plc our flnt cug l nl-imal i t c tool called LOVEDAY all that for ony nalytic purpose lt you can do that vlttun1t • co11p11't r • we 'lo'lilUhi Hite vcr ' nuc h to havo you ln WI A• 4 lllll'ller of fact U ma liGNw re f t EZIIIZ a onr I s RIM M ' TOP Iii PL 86-36 50 USC 3605 f t ------------ -- EO 3 3b 3 - I • • •' I ' ' · plays for Joth •' poses Ho•hu Ma Ve tAlnJc · vor use I In • • lty control iM -an ll Sls pur• a ·1 of options avaUal lo -to that ls tooJ vill prove to Htt 'tt s tal ha s MO hours of vorlk i t abll Jt nonal for tbls kln d of job be··•------------------ _1' Y of e cmi• 11 so iry • ql CS Chat v ulcf nc ir Jqioui • blo vlt11 hard11are •• • 1hero b nothlna partlc11lul' l ' unlq ac about • the h 11 a N tor LOVEDAY A l°' of atundon has llocn glven to nlcl111 411 the•pat'ts l 0 CIC OD UC 't •SD to w 11th e -rtlcu •r Jbp bf l YHI· Ill as • sea Jn addUion J a iina tt i -o ono wtl-t ls lflll • • to us ald ou h -c bvo a nu-bor of stand-alono c uc or•'IIHtd ttevlcei •·• Thh procr a • • ind its roJatlYos - · •• a povortul analytic tool tor us oven tbo' lab it l•n•t now PGTh ps anythln9 t at lists 12 car or so in t hla business ls 1Dt'Yla1 a 'toal - e '11 ero are • 1$1 oUl•r euap1u thac could 9 41s el ar • dUf1e1 lt PX di o E • v sua u of -o not yet eam • re thin llfl 'Ont else ldlat tho ·h to Nally prowoto uselul u- chan • of inforut ion betveei 111 111 Ibid· aaeMne A• you c u • o uny of our problem hive m physial-sclence base an4 1ood 1apl1y tech n•• 01 I D ct Ila Tho pYO blea uSU 1117 bas IODO• thin1 to do wlth tho WllBOwns bc lna roJDtocl ln smo way t hat isn't obvi aua - - solutiw s w tho equnlon sets ue not untquc 'lhnrolore l J llonholrni you play •TOUNt with who-c you bcUne bo practlc al constraints 011 part of tho problcta und I you aat o truni uo•• cm1vor An lyat lovo to araue abom tho whdOII of the othor cuy•s constHint os you can laglnol An exHplo my help hoN••• nique• caa d stn ly holp an a111 111t to se rue tM nss like lhe d nusic s of o pl'Obl or a particular tltuatlon 111at my be the kL ' to Ms undeHtudlftl of what•s aotns Oft llt b111lt that IDltff auention atudy t'llou ht ona o irped• IIC t l'4fl •l O •rl l c JI hi• rca r91' iuqaple w•vo dOQl e tOll IO -orlr on color dhpl•ys u an 11td in sonb a daia flll• Nsultoll Ln S0110 Yer ' coiod'uJ do$lp s •- 100d or tle s arut 111111 lllqlap for sure -· lnac it isn't clear that l t l s a pGrtlcular-ly helpf ll tll 111 to do ot a enenl-purpose approach to a prcblo•• Pm f 't ons rr b ffl qc m pPi i bopo that tbo pncadlna COfJMOtl have clvun you IOaO fnli tn ccmputer 1ppllcation1 u the PffHnt tiao Now let•s Lool tt the The future 1n the eosputu field it hard to teo Cllla lbUl't aN Hill 111 tl e P'O th phase wllf H t6Cboloalcal uvuees can scrriCU • 11 affoc t llhat you can do While wo aro slttln1 · hon saa Jap anua nsoardtff •itb fta» putUq th• EJni•hlnl couehes on • papc 'llbich vi U ICOto • IIClf llllTket A plil tJoras do nn inow so flat It b laFort•t note tb u Cho ELINI' and tel try wsines doe$ aot drlvo the uiduttrlal arlet to appr lable futurt •t anr e t•t· We sillplf uso 1 available Olu1y thoil ioDO tTends• • • 1'e vUl COftt nue an already ea1sun1 tNncl -tnant tl o prep r4t ioD ol clata hi tho field S0Do of this 1'LU fw11 U011aU7 l Jt e aacu I ik• -tiae •• aiow clo hon OtM - pan• 111 Pnct to SDNI' further lbe dhldina line Nl ft process i11 1 Alld ltl Al y-1b 1ysi - fllcrrv an cl l ' three Uitl l t wtcll cltbe thts tttn h hme-July 76 • CRVPrOl 00 • Pago 5 EO 3 3b 3 EO 3 3b 6 VP iif'S- ------------PL 86- 36 50 USC 3605 It J Z 7 PEP SPSPZ s n s n • • EO 3 3b 3 PL 86-36 50 USC 3605 •• ' '8F l CRl f Cont'buect presainie to duce oyerua pa• couse u• t o ant •to -tarioty of f'unctLons •• rrati• ll ltqmitt 11 • There' ls a doti tM timty rcmllts •• oxi ortaental t lllnsz s nce4 to- be unclernooct b1 foro they oni opariatie11n•I liu • Techftoio11 CJChts· to do it -- both In tho OJIPUtor flol and in the coc uanlcaUonsi uppon' field • s 11 pne'JWI-Furpose coaputeH vil 1 be• tho Perhaps that' stretcMna l bu1 not by 111 1 bHcanwbilc 011r folb 11rc nill looUns aUq liqly beart of tflos tfllnp wheNVor J»SSible f ' Jine Wo produce those Uncs ltlt gmt cut1 Im llpt Cial•PW Os e n uoptc ee uors NY bo i-eq11lrejl that •s mt thor u11 I thinl that ° c an lloYelop in the teleeottr tlutin s but I personally • cO iputttr-bascd tyi lftls that vUl diTCCt oui-utcn doubt Lt bee usa lt Nd 'e be •Y to ptolNl that t o o rdcn thin t a they oro reall econ ul n the • nunbors a 1 11a1 Wf l11 1M 1 1 a 1 11 1 -i1o 11 11 1 a t 11 11 1 l uire CQui-$e ts rt y ore tor 11111 uay reflect lbclck lnu» the procouina s1stu 1n a wy th tt 4 will grutl ' Tedui c sot-up tlllO It could bi that thl s ls A airy well Ny AAAl 'lts h'O IUM•C OOl at be t On• of doosn • t Uftdorttand tho 9roblC11 1 1D udous lUrtl who lt 1$ 11• s vu l peni t us hei'e 1111torlal chat to conctnr nte on t e ncplT • l•rc• -oTl factors '° netno to r tu•rmy omm Now or tho prollised l k ot the otec ronl c• A • t tond is t t one•of•a•Jtlnd things operation• area llhlch I c laia has a -- such •• LOVEDi Y •• will cll•t at S1P8lO Joco- carfno larco influonco on mt wo do of tlon s k t lho 11J1aly11u Jobs U ' bl'II pctl'offltld on gonoraliatd ollADp as 11111 ahow what I lfleaD the11 by people ll'OII o r orpi1tutions 'Ille SUpposo J'OU are In an odVa icad airenft flyanal 'tis pTOblca -Ii d isu it uted tho 1lllt lls are ing in a i ostUo anvlroftn lltlt You ctn lleu all distributed but tho catienslvc tool• cilst Unds of olactronlc tnnsmlstion ndu bta• nost 1y 11t NSA Pnt of lbc solution is a sooblcons oom unlcotlons -- 'OU MM it SOiie of sclenco pN ble111 How dO you con •lnco e11eryone these transaJsslons r pNScnt N41 threats to that lt ls okay to do wf' al c oz am csa Sa ' both 'Giff 'IUS$10n t Uld your Ufa IUt lltllCb But there 1S ar10ther patt How do tto transfer cnest And hat con 'Ou do alla-Jt it 'loa t ago tho data Uotmd1 Ko dD wo StoN ttf Holl do we p about retrieving h from ftmte locatlons1 you vere tucty Jf you could Just 11otlce tlult wo uc W l'Unr no11 on a Hmlted-acces data lleso you were bel• tracked by• surt1eo•to-atr alHila unt t lut th C l Utei- Mt perrdthd a wh1 h can be a aSlab e ·troll• few selected lo• t- »ol' 0£ other options to be opened up Nov it c•tlcns TM1 ff11f lead to rcmto oporat lon of u within the roala of posslbUUy t o sort out SOM ot the analyUc ciq lipmant u well 'The tho th at 11st•s £toa the otber trusmss ons trtado-of'f ls coat Rlallt now it 1001111 cheapor pemaps ldtntlfy then to lndlvJdwi 1 site deto send the data around but u e cauld change CGUP• that G nl l WCI' A Vfn' l'Cel nalyt lc tronll la tov n·d die uao of sal 1 desk-type ucblnCI uyb6 cv n cal led i alculuors by so• fot jobs which onco 111aro pcrfomod by larao uch ncs llo hovcin•t seen ellc end of this troncl y•t Cost lan•t the 01111 · factor hero tla can nnl o tho analyst aoro of• ficlont if bo cu chooflO hb tools 1111 sure wo can conUnuo to bo atilo to HY th11t 0 wo•vo sot tile largest cOllpUtU co plcx ln tbo f1'C0 world here at »51 '' -- l that • laponant A re11l pss •• ncn Y Ot A trc nd •· i S tbo 11utomt1c a nal 'aJs of ce 1 tina dsta Perhaps thlJ can •bost he e 1 · ed b the £011ovln tOl lllno thoi r c urrcmt threat to ou per-haps II ' Jooleing lllltoaUcall ' ot data•s pal transds tions alona vltll todars autGD tkaUy sot your Jemtna modes rand eua10 the •or Ti qulffCl ents for 'OU1' jmaen while ln Further 'Ou mt ho able to select your Uctkal U 'ICt hOII this slgna1 cnviJ011n c nt •nd assisn our vea• pons The voi ioas the solves m ' even b9 a uided • in SOIIII CIISH br 1nforat ion d tl ' obtained fl'OJI tbo sou siaru11 cnvl mentl or tab another eD111Plo SUpposo rou a1'a to • suburlne -· ipln lo a hostllo cmvlniiilont lhen you put your oalt up you arc Yfn'Y lnurostod tn lmowlna quietly and accunt•ly what Is around ou You'll usa llll ' son sor Inputs you • '° · can ot particullll'I ' lr thoy arc paaslve -optics IR acoustics and •hat wo lliJbt call l Lltll' if Lt wasn•t alnldy called Blectronlc S11pport Measures lbat do wo hoart Min is itt IJ it B tRffatT noro Is ltt llqulpmenU ox bt now and othc'rs a«-e under d - o opaent vhlcb search the r ulio frci ue r Juno-July 76 • CRYProLOG • Paco 6 'RJP HRliif 1111W 5 115 221179P SlllRllil O 0511 tt --- -----'L----'---'- - - -· - · - - - ·- ____ _____________________ I Non - Responsive onvil'Ofllleftt for slgnats autoatl Uy identify th1tm c akc dcuiled ncasure•onts coapa c thosa to Oll•board dnta bases and so•etlnes oven u ll th ptiin not only what class of shlp is out thoro how flar aWJ1y 1111d 1Xn ia9 ln vbat dl rcctton but ah vlun i t S • st Ukel ' hull nuDIJcr ill In aocond ond wi t h rcasoniJbh cOl'lfidence All tb11 l possible be se of opotational op• I plication of coiq uters and 11011eulcroproc •ssor s This i ort of opon i onal trend is liN l l estnb llshed today It h s a Lona way to go but Electronic 1'ai-fai-o b boi ng rocogJ'llZed perllo a bit late H 4 m Jor part ot aodorn forco nrueturo S e t coi putCtf'•technology poopJo have Dado thLs poul ble • · au sorts of new toctlc arc Nlw poAsiblo •- you howo m alanlfi• n ly incrGa C I iet at your roe sn of opt l ofts you can 150 to But tho o thlng only work i • You Uy know how the threat worU • Yau know toon ellQUgh to havo tho ul pacint l'COdy when you need i t and • You n keep it uparad so th11t its life• Uir c if lon enough to be pract lcal lhoso requhements put NAl teeth i nto tho tcchnicml ictol llgenco bu§lqa11 All thoso nico co q uters in the f i eld won't be 'Orth anythi n1 need If thoy can•t be told lilliat to do o to catch thinp r rllor In tile developc icnt cycle iot crgi lal l leQ$U ta ru11 tosts porhaps ae 4vo in order first to dotemlne ac curately how thla15 110rk and second to koep thcso dcscrlptlons currant particularly in con- flict s tustlons All of these tbinp feed back lnto tho collect ion prooe 1111 • -1 ls act1vlt le1 whlth aN also nlUt around c nputert It could bo that tho casputor b not a tool lD tho IU ltff Md Teltt flry buslnen t could be tti t it 1 th thl114 which acwsllr dotor- •inos td at we do le - sy reall ' wrk boaa _• of It if not fer 1t ff In su ca rr than wo caD say that c oaput r• are not Just useful - - they'N essotlall They aro horo to sta ' in II bl a va y in our businctH we contt gat mlong li'lthout thera And ith the tn nc 1 tovafd dbtrlbutcd £unct1o t ud 111th the hlg E'W operAtlons arket tbey will cenalnly bo cnton noro hipo-rtant 1 i the- lutut'O Thank you ror J nvi lng ae to adc Tess CISI Perhaps IIIC c Ol'I ro1 Low thi Ill' wlth nnothor prc 1ci1taUon that 11tll ukc a noro in-deptll look at aacno of tho lndivldunl applications --·H I IT RR H • HWfSRlaJ F UUUU-• I Juno July 76 • caYPTOtoG • PDp 1 l11P i Aii 'I 'ltm r re aa nnm 2121n us 211121 l EO 3 3b 3 PL 86-36 50 USC 3605 NlillilflM ni ·'° k1 ffat w w quutum P-4EtoP'1 MB m ell l'hdtMr Wa·i obZ Jt tD emr aUtcm it OJ 1 hfCll't lottJna the mnhol' gei lnto tho bOdy • Nnd fr eu - 1 · · niat• t q u tc Jiov tbo quote 1oe• lbut t ho '0£ hb •Ttido t ho oditOf' f'eels lhat it riiJ it bo a dvU blc to take an e xplanator ' •ido trip pt0'ole11 H8IS •have been around at lcHt H n uke Certain that OYH 'OM kilo•• • ct ly 10111 811 HaD ot l wiild vmitun a pou •that wo here at tfSA havo collectively cxpcm4e4 lwhat tr111tliteratlon ud •llllat lt cons1donb1 ·norwu on oor piwlcm Well • Isn't tnuletlon cmd it lan•1 tl'lftS • 1cri'pUo Let• s take the tol 16¥in byp0thetica l than llalllot cud - - h It ' C S Chai every •ituatlon rou•ro Ni Ike You•·re 011 a tour of J fo 5 yaaM thc spoi tro ot·tnnslit i-•Uon raise It a » ty n ad anciw and all t h dec uion tho So 1et U'ftlon You'd Hlo io uet SODI lbtsalan alrls but your guldo -tctlcf your bus· •we llado in •tho iaji ·stlmttCu t o bo ''ftdiaload Ute • halfl FJnall ' one J'O I and covoroct•t rc btcd ' r•t· another tlac 1'c b ' I ·a1sc 110ro a•t ncrv eta llhy this prohtoii nfusc s· to t y sohlld dLffteult detei-a lne cou1a J suaacst bOIH vtt t i t 4i because INI tlffe not yot solved tti ttav1n i1ent about 10 years l960-lH9 'transcrib ia anotho guy Joa Qll880 to ozchlngo o £ev re111rt vi th a cute•laok11'J slri· standing lo line outsld of Lenin's Mau ol wi You don•t spoak Russian so Joe '-elpt out Haro aN • te11 ot th tnt rchanges Girl o ano rzlna and reportlftg lbasalan orl11l and lbovt 11 ea ra s field ud' Russlon £ iafflmr-ir- r'lln'l l t dlsC1 1Ssi01S or translitera- fu ty a Ar e • • t Uon end cot proltinctal ln an aucnpe of thls uriclo IEOllMJ'I• JOO It I Glmie IUUI lf311tS aay t Glrl i X lc to l t the -t ioiual contont lot bo1ln by ollGrlng A Glrl I lieu DOJJJ I' Taw l a Joet Sho aya her naae h TNlara fer• o- tl' ftSf'onlill8 tr ll l iaforsation frett one a l t to ano er ifhe pro iert Jes ot o good trar suuratloi tch9o a-re ilsat the schem lhWld rouin H tnrormdon about tl'lo Ol'LJ1114l at pottlblo and th1t it shoi ld lie eully learned and used t y tile p61'sont us ioa tho p b' wlu cheao- In patticula I vlll L Ulla vorr emti not nd '1lahly detlnttl ot traullteratloft ar d then oxplaln 1 q sozie eoncepu and prcpertlat ot tr1111Slltcrj lion 'h'aulhor i tlon as r 11lll be usi 111 tha term be oddro er dOI Jl A Kiko What did she IAY 1' Joo • Mlle •• Miko I don'e lltQ lllllt did she say1 I DOH vb da ll it m mt onJ Riasst n-to•EQ U th nan•lic Joe She to tnov your noao Clrl 1Cail JIii ii ll aBIITOR KoollB t Joo She 4nta to know how ve Ulte Moscow Just then tha pi ff - how• up and ahooa you •both baclt ont0 the bat Nllte Seo ou ' '1'11111n1° I HIDO H t111e saco pltcef On ttio tus 'W'OG wrl to don ln your bandy• flmiac•cotebool loranoalinautn YoubveH U t auditory • cult ' and retent l vfty I °¥ ah pzltatl fo J lz11ktib i ahaw oot1 •uh#ot 2'bllmrz K4lh PQ1m ua Joo Jooks over Mdla al sllolllder and uy t You're 1pelliftl eVff 'th na au wronal acl ts d He lffl' Jwt •July 76 • CRYPfflLDC • Paro 8 GOHFll Hf UL _ -- ·- • Sit HIP rime •• I JI SIU _____ __ _____ __ ' - - -- - ·- - -· 68NFIOMIAI 'll m pt Ullu # Kale to a IIOWt 1 Htm a owt S PW'a Kak tltlft u Julta c Hoclala7 Me1awtd1e b ck at tho oleWI man en111sh dlrec tlon llu erpretHIOD lied tran•lo tion •r• concemod 1111 tb 11t 111t tho utterance - Hike a114 the NSA nonUnlJ list c1on•t can a -ac 'a whbtar about 'Iii the Jtnpibt dffiYct the mean1111 or how hard it b to· uatu tlloac coaplicAted aorpholo1lc and syatactlc rule For euaple Joe dldft•t botllor w coU Niko that llo cllan ed Tamra•• llov does Mo1c ow plea so luolf to 'Ouft• to How do you Uto Noscow'l or A lot happenc4 dul'ina these brief lnterto O JIJll'Din that he 0 tnMlated the nae of chaftse llnguisti1 ally if not n a antic• lly the Russian cl t ' Eron 1 MDskva to EnaUsh spuklna Lot's extrapolato o erytbina into a ''Mo Al f'l1' H tho nor1Un11 1ist it comreprcsentatlon of tho ontiro Rus ilJIJ lam uqG cemed le•• as easy for t e tr•nslator to lett-hand slde ''Tanna and the entire tnnslato as it ls for tho t'rat1scTiber to tranEl' glls l la1l£Ullgo rl h •hMd slllo Mltc•tJ scribe · SO the oonllnguht feels that that thagld be tho e id ol tl Al probl e a • uKE lfhat then aro those other five •nows ·t o1n1 In tho cbutt 1'1ey do not deal •1th the nlnas of tho wrds nit only vida their representation ln printed Cora lfheft Mike c Ucd for aaapl• t record lho RuJSl II sentences bo wonted to reDHl er enow ''ltito lt5'' • he did SO Otbln1 • liar to 'llWISCJlimON ''Joe 11''· But lftltlad of CCNINl'-•tylo 'lrin$cri ption vdtinc dOIIII m1 papa the utun in the orl W·1 tanauace as •polled i tho orl1iul •l bot • ho t ried to racord the Ra asiaa sentences in LatiP let• tcrs Oft the lbasi$ of tlio llul lffl t n l111s iottad do11111 two 4DOS 11' ho- aotebook1 IIAbt 11 o En gt ult Qlo-W C'l1tls can be done scl tlf c ally wt when lt 1 1 done by aut urs illc Nilta lt uauallr 100111 wobd Vhen Joo looled ove Milo's shoulder and corr ted the aptllln1 arro11 Joe 4 ' _____ 1-bo wn wo'ro fiMUy hlttln1 pt 'dlttl enp1ed 1n TRANSLIT AATION Jbls Js 11b11t this article onco 110 ·g t to lt ls all about Shen Nlko a h4 1'11hat dld a ho aay'l and the spol l tng of Russian vonls in LI tin lotten Joo • ntweNd 11She Hid • 11111• u eql 0n the bub of thoil cwlainal Cy 'H He arrow 11Joe '11 ' that dldn 1 t bolp Nlto m1cb 4PfU 1g The pffl1 lffl Ls ''Who' a JOl the one It was still ln Russlant Idle tMs p-t0cess oecu1'9 In tho CIONJNT lluslness tlult ls lis ten• tnie syatmT Joo u•ed a 1atC111 of l slltcma Uoo in which the IIUtslo lotter I is FO· Jng o people ta11t1ng i n Russian tllen puttloa dot -n on paper tn Ruaoian llt At thoy said h Is cnt•d b ' 1·'• If he ud ctsod tho NSA •r•t ho would havo vrltten j • t poUdi 11 Clllled TRANSCRIPTION 'nle process l s concemed suro onough w1t11 whi t the q eal er tldd non-HSAtts S4 ' that •H those '' j •s look funny bls esact words as he spoke tha but l doesn•t help tho nonllns uls t analyst ny Ho When Taun wroto the two nuo• m lot notettlll doesn't know llhat lt uans l Incidental - book anoi T aan '51• sho 1'UIISCRl81 D tha 1 ly OYen h this •tep dOO R t yiold tho IMWA $ into Cyrillic on tho baaia of tholr En£lisb nlna the voico tr4ftSCtibor has a Sho wrou Mab ancl - SIie 110Uld ha fe h11 nl Job to d1 ana ho certainly bU eo Ileen vtenl vro t HTaaft 1611 to t ruslhol'ate hat tho person l• talkill1 about bcforo ho can thtte ucotdbl 10 their Urd1t a J Ur g tra scribo It and lo that would aako th• pronounceWhat Joa d l4 afcor his flnt Uttlo ''Joto able ln 11tus tan ·ti ••Moek- 'oah ' and ''Yoh-yeaJln Rlcllc ulous hn•t it' Md 'Ot Cher oro EAaUsh WU to INHIIPR8T rcn- Mike OM Tuar11 81TOII words ln the Rutslan lanpaio tthich ue spell acl t•Joo f2t• lntnp -cUU on h tunr latton in Cyrllllc •d pronouncd ln Rustl llll in suc 11 usuaJly hack and forth fros one trpaMn 1an• m ridiculous 114t QWIIO to uodun- tt to ln Cho taMINl' 1Masi ncss nrely •n Invol ved in 111terprot adon t t Mita were OYOJ' io rcN Tu ira' a notebool 1 pc u cJl Jutud they iaro usually SnvoJVl d i n TRANSU • T 10 tho transformtlon ot text in ono mttcnt l1nguago to another vrltteo IGD£Ua10 • • ctnd u11uolly NSA tnn1btor1 1pec i11J iio in ihe- into and laboriously trzmsU ctTne tbt two nues i ni o r atln cllanct crts 11 °1r ''fllko t7'' he would o Uin • yk'' ftMIJl'' and DlhO Would ho NCO his own MDt 01' JGC 1 S'1' Or MNtld J ne-July 76 • CRYPTOl OC • Pago t lilllFID NllM IH411 • L 11111 SlitPIW C TRY EO 3 3b 3 GOHfilll'ltf Ikt PL 86-36 50 ho be ccm plctcly Ui'lawarc that he h•d e up opinst ono of tho tran114tor• biucst pnb• las O l' allOS TOblo l the rondlt ion of non•lbJn iln perMn ll lntMra• r tl ra t tonai C'4• 1' I i f Stan• qrap1fl t1 Colidal do 4'1bitiwlJ A A A Ba iN place naes otc not b ' nan sUtcrnton but by Nstodnc them to their e r i nal Lotln spoJUnas not NU on 11 'lh no but 10JUxon apdno 'UNI aq azbu 0 and o i - Ni znual 4 Ii B ie 'lle uual '' TronslUeraUon oa we will db cuH It then is a very spocific opcncl cin tbt n on foll'D4• tlon of Russlian to t in C r1'illi¢ characters into cho Latin nlphabGt a col'dlq co a spcclflc profOTrecl scheae Tllo tnn1Utor4tion procos• doos not produce a tr• nslat10f'I but c a1Tios tnco tho Lotjn roprcsontatlcm all tho arN rPti• cal inf ti on present In tho Cyrillic text Hcnco tho amntns content of tho Lattn-trnn llt• entton tc t ls ldontlcnl vith tho De nlna con• t cnt ot tho original Cy1-l l1ic ten and both tOlltS 'liO lld yiold tho ldontical translation 5evt'Nl Russlaft-to- Lllltln tnnsli teratlon sc s have beeit dcwe1o ed Tal lo 1 sbovs only a f of th The lUt of sdl02e b not Lnt d- d U be estaaust i e btit -rather w shaw tho ••fl vor'' of the trcan l iteration wtnld I r ti iii E % I J IC I N a 0 0 p s u' 0 •1 14 out of tho JS letters I tho Rus•Lan ll f r I q Note that • o 1cept for tllo ea ma Mone 1ystC1D •hlch uses non-letters tor cercai chanetora st cllo es uso CODblnatlo 11 of Latl n z 2H I I t £ E 2 z I J V IC L IC L l N0 0 II sT s N a 0 T u C f Of TS CM u 'TS H i SM 94 J II y X 8 T I u II X I z y G p R d • a E E B p R m 'I • p p D I I L I M i D ZN V 8 V 't• t 1 a 8 y •I B TIie •adout schmo nfora to • peater or es er degree to t la l ' T h of Ktur c ' ttiat -1 r on 11bi1Ul tr mnd ease of ut e Of p srt i culor intere t b t ho YOTiCMII 1- lin of th C rilllc lOLtet J e S 1 1 Z Q lit 111 1 l 2- J e iO R olrfl b - V CH SHCH y s T C C n y I YU ' zu f JU JA characters to reptt$ent cert aln 1llic YA IA R chuoctcrrs that 11 thP Cyrll1lc •11 d h' l dlflcd system wlllch oaiu the diaLatJn characters do not ccm1titute a crltlcs uso4 tn tile pnfotttd LC sy ten one-to-one nappinl • except for the unuaJ Morse synn aost 'E or 11tien pNcedcd by ccmsonant Othtr• sehonos contain lnhoNnt oablguitto In wbe iniU l or whm ncodcd 1 ' vowal that certoin cOftblnatlons of Latin letters YE or YE un Topte t txo dltfe L CyrilJit situuioqs aQd it h s i11as laposslble __ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __ t o Lell froa context lfhieh of the tllO poHlble Cyrillic sltwiUcnt had oceurtod in tho orl1inal tc t Por Any mothoaJLi lAns 1tlll readina they aro not •n on'°'' a pPJns The question Nhtc should oriso •t thit poinc b 1'1fhyf11 Why t ransUtc Gtct Or• lfl you 11Ust tltcn why ln se any difft rcnt v ay f Or ovon why vony iabout thct fact that Jo likes tho unual Norso 1yett1m whllo T0 11 as note cbovo teop tholr data ln dlffoNlftt trmtslitonUon schcacs thon tboy ahait be tulced prociuco a Joint roport or eiso 110eause ot th b own P tivo •bslons tbo · •hatt ftlNl it ab solutoly necessary to aorao tw o or aore of tho c0111pUtor here •• at a files lnvolYcd Tho answ r to tbe otlmT' qu atton 1• nolthu a c1oau1oru l utable but l s I kUevc depaadcmt cpon tllo awllabllit ' of 1ourcn 1111 osf and Han-y c1a1dcu1y SUIJ l' ndS 0 nM BG t • ve ha•o biad a tin-only alphabets cm our Tl io answor to tbo st plcn ot tlle Mi quttstioH tn evrit91's prJntus c oJIPUter cbal'actH' b dear and b probably adG1'lvatlvoorcol Olla o ltlrph r'a Lav 111t Ton Joo ond Harry sata k h ipacTJt ote an'• Jun •July 76 • CRYPTOl 00 • Paso 10 l PL 86-36 50 USC 360 5 illlfiillitli lM wm c rm m e n trnz sre azx use 3605 -----EO 3 3b 3 PL 86- 36 5 0 USC 3605 18Hflllltl1AL ---------- 11 - - • ioun uiva16 1 11ere au tho other· · • · · · _ - - - - - - - - - - - - - - - - - - - - - f'oUow lutd wheia we lioimd it necj SSllr ' t o exchal £e ddt il d • $OIEC iadivi dwaJ s· llavc preferred 1110rli ins wlth Russian rut•tal ln the Latl ft al pha• bet- mtlter t 'uin 1n CvrUUc l'bo lapUcati on horc 111 t h•t oqlJjpmnt -- a r# C01911ter proccisslna l ocd no longor n pN sont o Ollulitea t or j uHlrl cati on to t ran11lt onte Thl s tiowcver opens a WbOl e new Mtt of ques• ticms • • Do vc tl lJ '18 ·c • n ' to tran slhe ratef • linl 'f • llhenr a ncl the stil l enotl on 1111 ' 1oad1 d q uOJ t ion l • Idly not Jutt tmsl itcntc titt ro and solve the le probl Mr anS11GI'$ to the qud tlorus t 'OUld bo rcspcc lvoly Y Btt 11 use 5o letian •• a'ld 'lb 91 cep the data unatablguous and u savo riono1 '' To xpand OJ'I • that• little lot's con ldor ho quostloft '' Chy not tn slltetate every thinat The reasons for n J trarisl hontlna can 110 St lD U'iud ln 4tatllne £on as toJ1011S ra wral u•e J£ t1- mtt riak I • Nost l inplsu p fer Lo U le Cyrillic After al l tho fh s1i sr1 l• n-ac it wTitton in i hat alphabn and ihllt wn tho alphabet thet hod to loam in ordor to learn b ssl nn • • Jofost t lnsuls l c lio kina s id• dlctiooorlo 4tlaJC$ encyc lor cdl •• otc or In CyrJllle Mhothor- s ovlot-produccd or U S - producod • Cyrl lllc toxt • not anlJl auous if a The ans•crs to the oihoT thrto questions J bcliovo will pnvido the nol An illor to tho transl 1tent ion problezi Evon given tbo authori ty one wuld be hsrd prcsaccl to uko an ••alva s neve • t't'lnll li e-ntlon pol icy •that eervell all 11'0 daia ncadl of usors vary 111ldo1r d•t• bases are _ c oauy to aenorno and • alntala - d •ro dlfEleult and costly to con V Tt Lac la9 a c taJ a11 1t J difficult to preclic what 1AH5 be • of a data tile oae t110 or five 1u• rs after lts inception 11 et qw ti oll t which l'ust be ul ecl then an • INM1 • • what ls tho llkolihood th t thla Ille wlll have to aec pf tnns itorat od dau as input • snmt GEP •· wtlat la thtt llkelihoocf lbat thl• fllo will bo vo t o bo nixed•• ltfi tUe Jn a ond tent tnuu1llt•Ntt d fom or I otbor tnns11 en'to4 for u I poraont a nue i s 5J Gl1Cld whh o • TS t there ls no acb11i 1t - wltri ntml s pcu cd vlth c • TS • 1t'RT • tennlnat IMP• lnta 1C'414tO Hcssa1 Procouor of the AJPA or PLATF'OllM natWOTk HOST • a centnl cogpucor ur a ain procouor ' t • 1 rne July 76 • CllYP'JOl 00 • Pago ll 68IIFIBEUHltL rrmns r n zn s e rru a • 11 I I EO 3 3b 3 PL 86-36 50 USC 3605 GINlillBIIAL outfun -· vhat 1• the 1 U oli hOod that tr nalitoratod output will bo roqubtd IN JI rhls f'ilO in ono or sioro traulit r ttOcl 1chcracst ' t • a use o t s ta • not Ungulttlc- no 111Cfttlon of transcl'l H or tl'III S• lator and ls very l llQ •c ntlo 1 of clai e xch ftao or other offices lt would 11ea that' our choice to tnnslitent• b th most cost - · offocdvc Had the lee JOH 'the other way ' soxo ccmreoti au 11Wld llave Ileen aecesuey - · poralt Input froa tfals 1 atln Dn1 ' device • Thoso conventi011• 11011ld him n qulred very cloao scrutiny to inS n'e us abllltX' and 1cceptanco b ' tlloao using tho circ ui t C SC pc D 5 • 5 OU ' 0 ap• parent th t char ictor fom ro qulTt'lllmts should And con II• ca funcilon of input 01' 011tp lt ro• qulrmtenta od thtlt U can lo 1 csolvcdtt 110TC or less dy 111taic-ftl1y at t-'hc usn•• option A wh h aost ot hor d 'l'IPJ ot cti on1cablc proarDM then is cost for thls a cost BHuncl in U' 'IU of oddcd complcxlt y of Jn'Oll'llb and run ti11c a tchllc these coau can be los onad by 111t'1ndonthatf on 11no opti• hotlon of tho ''tl'us- Hterate'• and un-tnnallterato routines DH mrtt constdcr •11 threo functions •• I 'l'Ut S'IORAGE ond OI Z'PIR •· to clccldo how to handle 1ny at vtn filo Consider Uto followina t o cnsH June-July 76 • CRYPTOLOG • PABe 12 IIIFllifRilAk MHfBI 5 Hit mr r AUPRIEI 7 are 1 · I ·-·-- - I PL 86-36 50 use 3605 ' EO 3 3b 3 MFIIOIFIIIJ PL 86-36 50 USC 3605 • OIIIIIEMTI AB IEOI EITID •ttre •i the Al-21 IATI D 'Hap WH coi iM ' ·I M tho IW'llll or 1ft SOM case on a se ara e typewri r ne operator does noi hm tbo- Pm fl-ibUity td th the AG•22 IATi so ho Jmst prov cle COGtilllaOUS copy Two data 'lnuai cme aftd 011e tor eha tec would 'clea11 f t up 1m1e but can we 11fford' I WO ro ina dovi o•' fen OJIO oporatlDT'f And ow aboiatl 1' tNffic were •wfs a10111 arq• no£ •i I 1 0 ttta AH in a4minutreff v• tr·•fficf They ore more Ute m •••1a• 'chatter ' A• rot ta rln1 requlraants these 1hcnald be -1il lJlm1 becaua fNOry cl lod 1'tqulra• kept u a d4 tnctilon tncrcsuos tht hHc 11 thn lq»ftant• tranwnlona vill bo nssc II bt the oper•'to lt - re ta11 aH added tboy slu utd bo i tade tloul 11hlch would allow for 4t fforOff90t tn 1ntncctpt Of Orator c apabllltles •r d spea dirficulty of tho t • rpt ltfflt Ol' · lun Jr Ide ___ _ _ _ _ _ _ _ _ _ __ Ir •• In hh 11rrticl• Cec 11 PIIUU a strona Of' 11c thrco r otencial soJuUOfts pre tented pitch for altqt ln tho HF' Hort copylng aethods 't y Mr PhUUPI for iapniv na the AG-lZ IA'IS u a solution co the Wllu«ui ful fforu nSJ lts the aojor -- hub should be on tM autg111 1 • Sb• xtnctlon of t _ I l'Olll ue thtOUUH O xe n rxw t __ le dodci s and softlr3ro tTH I c Oflp lte1' s omt i e anal 'tle edltlna tlltoup use or tonninal Certainly ° etiould aet u taTf'' cntlt ' was uHd in tho 1960 tests of tho f r t AG-22-tn o de• He did not state 1ehlcb i • t» ai I havo al ay i en 1 -- 4 • that it wa1 tho vo y storeotypcd trdflc lrttcwcopt If t-rue that IIOta oxp a n - by tho tHU were so aucens£1d • ud vhy Oo frobl - t olna encoimtored toda J' are not wltb that typo of ttaffic but with otllu Non• tar• aou which havo 1ulJscquentlybeen covend vltll AG•Ja lA'IS eqvtpnc tt A propo cd tOlttcion hid placoa tile bvcten on the it tCRi J t opontor ver ovoTloob tho 1Wt don• of oven tl o bost of operator• No •tur how IIIUCII a Ni no intercept opera• tor- 1 h t about bls taraet ho - 11 atill pot bo ablo IIO COPY wory hr bchlncl the cha1'11C• ter Nina uansuttcd ospec lally when unlntolllaibl• or unusual infiol'Mtloo Is belaa lnterccpuxl He cannot tile the tillD to think v ry ni ch allout what ls bcilla said in o 1tT1na of chatter or he vlll ai•s part of Lt The 1uaaosU011 t-or ''swm rhint' 'by • Mol-H O Wlrator copying a live tar1et eunot bo cClllpUd vlth eisdn111 by a volc e umu1c rlbor who e opiet frDA a tope ldlich he can stop at any tlae 1 ave• tbt then h n0 aHd f- 1' the Morse opera• tor to COP ' eve 1 t N ve-rfion or anything llrl4 ho dldn't in the dayi boforo tho coaputn tdlen 1 only COJ1«rn was prod11cia1 • pa o print Ho slcrply tacltH up the canl•a• of bb typf'n'iter and spaced aJonc until ho deteoted a diAll o then ti pc6 it ln abovo t o otl1lnal veut on Maar ases - ere a11 ay1 foftllltted non• repotilious c bsttcrr set1t in tbo 11lddle of • qtich help posstbl• fro11 the intercept open tor ncl 1houltt deR lot aoftiatro io automtlcally 11 uttatt •• nuchas posstlll•whhin lll'llcal ' • •hi • Ji H• VA t11111 t P1J h11 _ that Jor - ounessH oabt ln liotb ot uese appn a u whlr h vi U pt eYent our - wor pt Oduc• 1111 a vert aood tlAta base l'roa thoso oeus alone We ll lst lnje t hullln lniet11entloJ1 into tho AC·22 JATS convertlo pr0CO$S belon out f01111ttod acLivilf data beaq which an al• l't pooi- t ocou IIOrtbless A clual-Sct'ecn teftllnal fflftJellODt COUid be used f cn this 0110 ahoidna the orl1lnal version of tho AG-22 IATS Intercept tlld tba other 1hollln tlae clau extracted JIil lomaitcd by th• en puur An ualy st •old Judie tho accuracy arut COJiplcteness of tho co puter roco abll insert c11an1es adcthlons as deslnd 11 perforala1 t loso usb rigu ar y i •11Jd pln ntwte lRowledte Of hll ••ti Qed tar1et1 and IIOUld bo bonn- •blc to •icct 111i111ifi °t c b11Psn than an software ptck•ae -e tan over develop ulntaln 111 CCIIIP'lter would sw o bost £or rapid con-olatlon of 1tol'OIS data fZ aa vaT1oos appn chH capocuU1 aver a lODI par Lad TheTt is no 40 lbt that the best possible uth·l• iy data bale 1'011ld be built for b proces iqa 111• 1t andard •r•ent - alnst using analysu for thu b r avaUabUlty an Ol' UfAnt that but t t b uuaen alvqa I all bac k on It noce s ry • can mb tho - i owe r avallablc by ctwiatr111d11lon dlnctlona ••••• by H• duclna or oJ'hdnatina SICINT efforu aplu1 countrl a wtt1ch allow frffdoa of movement bY GONFID ltJ IM liPUIH Plf 111 • 11 llidllll I I 1 I -- I l - ·-·•-' -t •• @81ff'IDMIAL £orcltnert whhin dteir bordan ad ahlft lng theso l CISourees to t-flol'ts ogain t tho tlptly cot trollcd COUfttr1Ct9 otpoc lally tho Comumi st or m 'be CtVen by uling somo of tho prolJl'UZOl' blllots that uo boln a waa on tho perpetual unsuccos• ful cobVQJ'81on Qffons Such inaJor dcclsio11s nay not lie l'Oquittd thour b u 1 u not ccnvlnced that the ouilysu todoy are as ovcnror ecl 15 tllol r t ouo• IIOUld have us boUovo If ll Ull 'IU don't build Ol' COJ'Net dat4 files llhat do thoy do '1th th11h tlmcf Analy• sis and reporting takol vory little uno lf you have good data bases hc11 vlllcl to wo'l'kl I 4Jl 11fraid we 11ro palnt1na ourseh'os into a 'or as wo Joolt aoro and mire o the COlqlUtcf'll 1LO holp solvD our real idglned Dllb na pl'Gb• tea Unfonunlltel ' the old Sa tiftSJ G tl't• in ptbage 0111 t nUl applies but vtio cleans up mchlno flies these dAys -· especlaJly utl 1ty files Many ualysts don t won or O r01 soc t raffl any sioro -- they uso solecUve f'Oflllltted uchlae lhdngs and I vonder Just wbal kind ot tcchftlCal cducatlCft thct 1 TO tOt• t lllJ tl'Om vorkingidtll all oE tb6t sartiage u prodlll t ls bound to su teT as the aoro e c• pcrlc iccd ano lysu departl A rin u l c ox tn also ncolla to bo mdo • bout nnother •tpect of the AG-22 t 1'S pJ'O Tlll1 which is ¥oey lupoT t ln tho 111 UlOSoilGflt of inc er• capt 1ite1 With tho hl ft-1pecd olcc triul data ie11t o capabi1lty o bivo I Oday the AG-22 JATS oper3tor•s c loto lnton opt r a h111 qtlic'lcly tnnsmiucd Mick to NSA for PTO• ce sins and analyst TECSllf preparAtlcm ot tho 5lto whicili • • L itiatcd ycirs go to aJ• 1011 for Uwely lecuitlll ft l WardJn1 of intcrcepi l I• and latDT mitered to fCHld P l er c11n be c UDinatocl Field proparation of 'RCS lb ha» t1 l'Vcr bttn Of'I good OT1yflo111 ba IW$o tllc job has Ull lly b«n avoided by all but tho leut qua11£icd eiwlysrs asslgnod Unfol tunatoly Jone after 1asullaUon and 1ctivat1on o AG-22 IATS ociuip 111cs t on uny tu ot•• pnpan· 1ion ls tlll beina Tequlrcd of tho intercept sites tho rcasr m bcl Ins th t wo 11tU l don• fl lvo the tofiwaro to do the job and aro unal lc 11111• vllllng to absorb t duties here Appal'Q'ltb ioo 111u1y of our mnaacrs ore llvins in a dl'eut world roaal'dlnt AG·l2 1ATS II lid tho tlno for 111•dcnin9 ls tong ovol duo Lc t • s set the analysts in1101ved in cJcaiing Ill' thi 11cs boforo their tc chnlc11 cop11blUtr to ay tia '• tekn 11 1n7 incmt to becoao ncmlilall ' PNCf U ve on • Morie posiUo i uncial' the instt UC• t i ons 'COpy ovorytbins you hur Ho or sbo vu 1 by the t1l6 of that flru tour have bep t o develop som nc Uon about what tho targot b rdU ' dolna bllt by tt at tiacl he 01' lho 19 re•dY eid or to pt out of th• service or to ncnlht thus tamia L o n to soile other job 1'11• collecto r RUI' tho elld of hh tour alaht be exporieaced enou fl to harldle ao so• phlsticated IUtructlOftS lndeod ldpt 'lCOIIC then but he vill as surely as 11 ah'- foUoo cllly be replaced by another bealnntt 210 will not bo ablo to handle 1110n sophtsttcated ln• structions Such cyclic infu1lon1 of inoiperlonco prob• ably nproscnt a lhr ttl11 l tactor lor- COi'U as °Jl u AG-2zs I thlnt that eott experienced operators would asr1 1c tllAt o ought to roly upon tho operator fol- oorc judpieftt IU'ul dccl• •lon nmct lona but mybo wo ought to tako a hard took at the lovol of incxperienco of tfle avorogo colloctor upon who» n have for so many 'CAI'• rolled TEIIIIIS lperatlon-Faalllty Sill O the 111tn polnll of CocU Pll lllips• IL-------------' 11rticlo la ttle recoaaciodatlon to seek the PL 86-36 50 use 3605 solution tor obtelnln1 ettei- re5ult• frn lG-22 IAfS data by cbafttinl tha n t M ••1'lf' •• copied t trJod this idea oa sou • r clviJian CaToor oporatot's hDTc tn the 1'E '1H S Operat lora Facility TOF Aa I waa i T la 11 ' HJpcmsiblc alons with I Jxk Croat Art Fctcho et al for iJi• dellelc s-ent aid procl1LCtion of the AG-23 DAta Pnpan tloal Set wbkh nplaced •u1e USAV 0311 J fcol fl acldtat roaponslblo for tho sins that n dctlpJCJ's have pushod on the opottt01'1 and bope to ailo oaonds for past al stale• Tho follovi111 or- unodlted caatents fro i SO of tho best perllOmKll ln 0ur OpC NtioD 11 911' CClll ICIIU are oJland b ' 8 f i'w of DO' 01o111 Johll Capfl1 • TOI' r ------------------ Tb u Mr l'hllllps vould tN it tile • toms 14poor 4 11 1lty lf 1S output whb the vroiia cure A better output dop nda ultily on deaUna wlth tho collector as an indivldU 11 llDd Illa E• Bto •l• ' ' odure colli ng pi-oeed ltos or othet qua Uy ontYOl neasures ThO collecu r ts the f 'l'MT do so avaptn'3tes of -------1 needs Flr t reaction co Mt Phllljps' article that our entire system it 1ured to the SeHI of nexperic11co af tile coi'1e4 tor Wi th r lnor oxcopt1on1 thD Norn coUectot' l• a tlrst tero operllltor llflo hu m t noP 1lly en• countored l llniml JJO befoH• joinina the ser• vice who was c -ained by tho ' 0rvi« c alld t ho Ny zs EO 3 3b 3 EO 3 3b 6 PL 86-36 50 USC 3605 and Oftl ' uralnoll ' with proamml ia of tho SIGlfff world and ptodt lCD$ tho product that keaps the ttst of us eaployod Iran ' pco• plo ot Uil t AgeMf ofte11 lose ipt ot thh fact hoveve'r I could not accuse Mt PhUUp of thh Thus tho tanacmt ls uos Ito llCfflt iou en p't l bly ON re1ovant o th• cure thlln airt his nai11 points June-Jul ' 76 • CIYPTOt OC • 1-gc 14 roars mcrffi 1 JC 186 £1111 --- ass tubl -· · ------· _ • EO 3 3b 3 PL 86-36 50 USC 3 605 --aJNFJ tdlflAL In fll1 view the 4Uallty of the collector out• put my be tuffcring becausei • Collectors soll 'IQ to bo l oft out of U16 plannlftg cycle on oqulp11ent that 1dU affect uai Too otun soao u11 t aer ona1n- pro1ramer or an11lyst 111 de• cide lfhat b ben tor tl e coJlectlcm phaso of the SIGINT cyele wlth the result that the scheM c Hs not aoet coll ctlon neod • Tho eodbact cycle to the collector is very poor Often tcelu ical docusents will bo vrlttcn £or internal KSA conswq tion an 1'cl valu lqs to tho collectol' -ho nelklls to u o tho dtlel DOnt aosi Min ' tecbnicel SIGINT report TSils 41'e prlsio o 11tUPles • Tho current tnad 11 J D 1 zzutto t o ens'lll'O collector divcriity Althou h this hat benefit for c l'enlni lntorost 11alt ing annt ng • po1i cion easiel tnd al-Vi n • person o wider ronao of experience lt docs provcnt Ci n lnl'' ond waulz tna to a cleP'ee bias 1oing on £or iOllo du l e MOUld b when • linlc reput•J orrects or re at• ten or I Ochatter ·• the operator almply raJtc s a M to thd the Cl'fftio l of OJtP« TO 'lbu l the ln ettect to gist ii- to swmrho bo ivn4 thla depUI nuances o col1cct01' ' COUld IOflrn ofter point is not possible ldltn developtna new taroxpoaun to the blCIII no not rtallz ed cots or copyln1 err atic tarptJ r ww1n1 •rour toraot his • lOt to d o with As for dovolopln1 nw IC Ulpcitftt f tl6 sore the quaUt y of tho i ratcrctpt tap 11nd fjap th t tho 11peralaf' hat tO use • Thtro ls • c on tcnt influx of nw col loc u o operator sllould Ibo ca Uecl on tor hU vl a tol'J to the Ylrious lntoreep f it•• 1ht l r and Ideas 11119 ovorsicht can tio soen in tho tnlnlna h oftn lulphuatd SCICOC' dly ••onalnccrina dovc1opac1 1t of tbe 11 MMIS e11nse le irany lloctots wllv no lona-r tranlt io be Ono Shortc Gllln of tho 1 JIHI console ls tho collect ors aro forced to •t colloe• Md tho kc 'boar4 nolther of •1th 1' t lon Job This lndl atos • need £or NSA o o t l t h -ore job 4J v r•Aty in th• col centered within tbe Po•itinn 'l1lft h11t11dsf't Jaet tho console and n bo l c icm field Many only wtmt • tc pon ry Pfl tnides outwerd broken off b ' cine's body or chab Got to Gen• brnk froR coll ticn tlon tho 1l1rc off tho scope and tlle keybcMlnt A 1e 1 n nt os ect ot his article w1 u· tho operators WNI Ntod about r ore ua• and ch it ho scre as to bc loias to tho school of tbougM flap one would seo thU w opentor 1• • VffY whkh u1uae1 that ''data a•thorlna with acblne r••U l• lndlridual Fnm his past QOVlcdae s anlpulation will result Jn pntc uclng va ue•fl'et he aqy -ven ako t he a11Plys1 •s Ql aaler and lntol'lli 'ltlon •• He alludOJ to thb in hh tlrst po ltibly enUghttn ttlo c oaputor p op anor J a parosnpfl It is my iew thAt tico and tho not sayin1 that oe opontor b 1t1-l11tol l18 ible devoted resoucti by ll Sn socl•l s itn l•ts who but I A saylttg tb t 1t1 l about tlao that be tbou pt tbt ifica im1 mul dalnc q mipv b re phcd JloJ' wlt Jio Ut tho operator you latioft wuld produi e 'latue-frce lnfo11Dtion aro out ot a Jo'I ha i lar1oly disprD'ICP tb • t h ry Altlioup cay ''° laraer sepieftts ot lntors2tlon c bO pulothe oalysis 0£ tho dat• mid the need for the analysu ttnatn This ls one 1'e• son why COPES as a systMI ot collocticm U £ailing •• a lack of propu anlllysls o the 4at at harul tl d with SUbHCillCftt fcedbAc 1bis lad of proper anatysi s ls an uy affacUna tPiequaU tyoftho autpul of tho eql lac DT 1 bcl ic vc tMt if tho t bac t procedures uiprovcd th m so too MOuld t a qlilllity of •• ion - T•••o1r1th - Frank Hyland Needl•ss to several discipline• r1pr1sentod in this office SCIIIO of vhl h I 11'eotp0rate In thltreaponu llihlle I aot uniqaaely i Ul Uied t a ft-aao this response J havo llad the bonoflt of havin1 been 11n SCA Jfr r so COUeetor 8Jl NSA Traffic Arwlyat havins ta114ffl uniloraruuato and arad -io tralntn 111 aystem 1111a1yals oporattvns rosoa reb and prognaiitt1• Ill la roE'orenc o to d10 use of o MJC ary or eist to 1nurce c aanual •10ro tnns11ltsion This ii oncept is totally uMCCCptablo '° -rctHon c oUoctc ir• 11l lnt or ono ll'Jl or All unual Nots¢ tr11nBi si olls II copied on ••r Rr Phillips• artlclo prcmptcul mny COIIDl nts frva • of tile I havo now coa full•drcle and crivJUan MIWff J- •t PftjO'llt IIOJ'Uq as • loctor lq G-TENNIS iI An -rJn tho author's pOlnts ln orda - rill pnhaps 1- the boat fomt allowlna ttio reader tlon bo h piec sJde by std• At tq outset I rill -stat tt that he obJcc- Jun July 16 • CRYl'fl LOG • h te 15 COMFIBiJlilil Jl rs au ttn • r t su 11• r _ 4 I Oillillifttl JL uves of COl'IIS and IATS ton atUne u well as t hoil' prcidocouors alld contaponr-ici ai-e Durln1 this Pl OSCftt porlod or Halted 110notary resource in tent of an inpuC-pJ'OCCSS-OU t'' IIIOdel H5A aanagus kftOWll for yean afld h vo t1'lcd in • vain t o tell oU vr•• that the slpt fi •-• of a ••stedft Iles not in tM faci· 4dlliTable 4UalptJn EO 3 3b 3 PL 86-36 50 USC 3605 P'e to adjust co• rolat Jvcl ' unlm J ' Input fundir a by 1torin11 the process change and intorference 1ev11s ras• al'MI Call C ith tho annual aaJm n ru11 r all I bow No what wo cal the ttgood stuff ts fowul ln tllo seo a-• ln1Jy saG1Uost of doiaill noi foumi to glsts One u risalttllla opentoi- per- al au ln • tftllliinl to lcrftd bis traHlc des11H• lnterfON'IICe leveJ vhlch C8UH ot hcn to 10111 a JKo vacate the h•- · in oNer to axlalt e the 11wcp11t RnUsti• caJly of courso ln te1111 of pae theor un111pt19nt ts surehin1 for the process profile •hlcb •lnlnUet tho ckaa d' dOlle to t ho intol - u onc• •££on throuch Nducod ftmcU111 l•voia Wet lrnov tbat 111th a lesser mmbar of coUectoTa analysts and ropor •r uraye4 apinat • POWlq JWI Nf of ic atlon tem nau of incra1in1 co -ipl ity alld sophi•tieat loa we carmot realistically seek to pl'Ov de the sno or a hip lovat of lentico a ithout c hnqlq in 111 ' tho unnor ln which vo oporat e It we are- honost howver we vl ll noc v Clf COPES or IA1S •• havina been chts led into ftb I' •e ' slonc tablet We are nov onpJed in a 1oarch for the ••proper c ClllblnaUon of peap1o and aaehinos tho cwbinatlon vbi cb wil l oaJdnlzc tho A ency' • output under th• 1lven sec of Cundinc COJ 5tni ta COPES an IATS sre expedlrlCrlts SG ewhuo n r th• aidpol nt llop• fully on d c cont 1nuu11 roprcsontlni the cvolu• tion of this sen h tta rlna -- • that nau•nt thousta I 11USt al o say that In tcrn s o 11 EO•I attalnant--type IIIOdtl 1 find SOM of the ac Uvities vhl h NO carrild out ost onslHr to achieve tlle stated objectives of COPES a ut IATS pun Un io HY tho h Ht In order then t JWIH N JC1- PhUUp 1• nusinas 11d n y own COlallfflU • 2'M avtltoJo •tatoa tm tho qoi 11t _ '-''· ao tvn gl Up Md gNai '1p e tlaats ' fl uo r«zn tho o-au uoulcf JJavct n g wt OIi annZytria 11 u bu of tu NaCiiN Of tNffio COVl'IO No Al IUlllfflt floll ae Ol'l thh one Nol'o lator but $Uf• rice to say hero tht t tho desirability of this IN t ilCJ act •• b another utter th t is open for discussion • rNNt 4n7 nU ltw• Wl' 'l I ca fl'lfJl tld t t u d for oou i - t'M flu A P NOO't'tm q1rra• Ntwd to U y il-catJH f'ltOl I AO- ond I A'lS aro of N-V poor qrallq • i• •rsu• vlth thst b•int be dfl ietecl for yous with printouts Apin I cu•t to ' which lloTo little Tcseni l nc o orl1lnal request for lnfoma J on 1flta I do tato lssuo vlth tho JIJI lt tho Autbor 1 s solution At II laur polni in Che piece ho H 'S 11I C MJIA _ fM- ild olltlr a die Mnoao MEJIJing l Ol'SO P fttOCt W lua of Ol1PJIUIQ c11ur tlil '1 to t11B Uea of at g' u Ml Otlri' T 1 nQN Z t to 'glt # rg n uo oc CJOllr 'US'UN'JtiOtJ Mo1 M collectors hHe 0 j another han1s cm ho ends of qllfflCy I o tJuo lh• Mx - • - - vei6lai aon WWW it h ono in Ri fa k r bo'Ki Norao coll ton •um lliftl • cransal11lon hlehliahtlnu tho t r or• unt u cts only That boas tho quost l cft ''What b lspod ntf ' That can be o lar1e hsuo with the IIDDUJ t of autor o• ay joyed by a coUKtor And 11' prac- · den l t ls daonnr•tcd daily t a aentll 1ot ensues vhereln lhe threshold botll n no -1 ad 4dJeo 1i i rdsecl until it require• one bettor an abnorulity 'to 1nu11I• Girt of the 1ht c atosorr into the hl hH ht Tho l tuatiVl'I b noi• b•lped a all by he tnovi• tobl• inc •G•• l• t • ou q ut • cp•c mt l on1 of tho c ollcc tio t --- aont staff -- •ou 1lnce JCN'r• only na- oari lna y 10u ean bandlo thoso fhe extra c hDrcs • l di not- tcnoa c tat tAo goal ha• not boon JtOC1tu eL • • t1toM q also u llll1N oaru tu n lit ti-_ bta • • A good ·pos Tho suspl lon that tho cimputor proaras h vo not bcet1 as tlabtly tallorod tho liq ut Is •ISO 1'1aht on point The point OYor la tbat • coaput or pro1ru c im be os tlghtly tallored as i i noc1ssJ 7 'l'hh Sen ' To aal nl in bu1inoss beslc lly because of variation 1he v riiat lon in ttans iJ ttlnS equlpaont uso ot -ru1o pt'OCeduro and· arlation in lftlllvldi nal$' still aro tile wtana by tdllc h n dls Unguish botnaA nation•• sonicu Tadio poupa anll tranaltti11J opera ton lo fiut ha 'CI a tol loctor IIIUII wt a ponion o£ tile aea is by Mtllcb w analyat tDn diffC'l'entlato anti hen lntcrpo c A •chlne to turtbar sumarh• thl NlffUI IOo-5 to IIO CIJ bo U- thetic al to the •1oftCY'• stated pur- pose It is ono thins for tPLo lktput• aont of A lrltolt Uo to promto the 1rowin1 of tobac c o ithll HEif spe M s nillions of dollon oa cucor ttsezareh Juno-July 76 • CRYMOLOC • Pogc 16 68ilfliifl51AL uee·w s rr r seer n cnr e·s n c n r lllr i 1181' IAl EO 3 3b 3 PL 86-36 50 USC 360 5 t o apcnd th l J a ' p l ' l' l ll'ft0 ' ' ol l'Ta r J on - eq n'pnenl e 4 people to etAso tho ollcitcd vari•· tiuns doosn•t teem defeati blo to cw A6 ' A • eJlbff ot tho ort slnal asency ittco hich f'ot11Ul1tod tho SfR1DI pio1nu1 i ftd • • a traffic analyst I lh'ecl 'LO TUo it l ser1®$ly q u estion tho visdml of our i ddln1 to the VC II re 1 lll wp U e • ND ffP 1 ocalmit i t i tJ t v i t 10 cwaew ioo1c ac thl TlP M 1 H t ook o i• l'rt lal llhat E 1o cu r f COPHS a ot irt MS bfi n aule to- lmd c ie w t in U t cffitt on a epnf fitt or ta - - GOi bane Co1Tc ti o11 A s un hat been 1111 d• toward •• p wuati-uw UAJa of what S s i11port111t The dlffol'enco t bolicvo is a very slpifJcant ono J a a DCllMI' of the STRUM omdttea l ns •blo to R i ni SOllO lnvmlumblo con• ccsslons on ronat flealbiU ty froa the ui I vas away during t ho pedocl in lfhl cb t be 1ttcflS concept and COPES gorai111 1tc4 $ • collfftor t tla- o to conceatrlto on bol I copy a loa5t •• P U h o ften l'IOro Ill I concentrate 011 i AG• Zl lATS as woll H other SOftlfne ey1te1111 do • noi •llo• £or ert'Or• not c wen c•IIIT•WNIIY onoa f r tho 11111st coepet e it and ladu1tri_0111 oponton wu l occa1io any COl o to wort anct not foel quite q 10 pu aaot icmall or U there • mall subprosr- SOIIO• where 1n the AG-22 lATS S 'ttea _ lch vlll ac ount tor any en-oncou data tlow r l 'lltlng phy icauy froa thls clJle lslaf Also doos AG•22 JATS provide a sub roanm u hM cotiv•Uon f '11te old abp You can fool sone of th e peo plo SOl e of the tine IIUt not •H tho pcoplo au th• tiu 11 does aot apply to the AG--22 lffl proaraa•• tn su ry I bollovo that the Cllffent AG• 'l2 pnl8 11 can ylcld good NSUlt5 • • to a plati 11u or 90•9S per tnt of tho UJ1C But this depends 011 coapetc t sotlnted ojpcnton IATS John tJap I - Jlroa the Above comenu a nd others •••• board but To ur prlntable if ia obvJous co• that vo tiavo the baste l adiro of knowledae aq l• de ip • bottor s stea 11tls cadre bas t•_ppocl lhe TENNIS koytloard and OJIHII• tor's coai1110 was deslpcd uoro than lo year• after the Ml•U anci unfotwnatoly 1'0 0 of the lesson• 1eawtcd qrc trnnsleted to its no desip '1110 AG 22 to1boud arid tai lni system has aablo i o not bee111 beccaro Ill 11 stlUlila1'd and tbuero o I COft U'IICt •pccificadcm Unfor1 111Jue1y tl lOON ARCH£R if proccodiJla dOlfll tho ano path 1ho AC--22 ns tile tal' Ot IUolt l plt ' tile poor da a and Owor corc Inhibit proce duN for copy1na anaiysl$ a ffP6tton 11111D mst -'d t barred characters r• dovoloped loly as• thrcuth the s Chlrt6 S 11bol t and the CffprODlso lletiroen the mchlno OMlyn•• desire stilt« awbarcl fon at to pt co sr uaaarl c1 eopy • nd •11 the •hile t he · and tho • bariic11l rost ricticms of tho Tolot te 1Uuadon -rooa or CIA Mqs Ort t Ut other end of th U ne a st l n tio1' ''soroe dal41ih 0 11 aro no mny r»rc NU ru t o a oe tho 11uthor•s COIIIIOflts which could £4irlT bll nadv fltl1 ts enough of a surt Ho9etu11y •ia 1 who have be ln sayln s1lentl1 to thasel e ''LOtd cletlval' 111t fto11 tile coqiuter1• 1 wlll J o ln tho gr° 1ng nwrhcf' whD u Yi ' l t C711t 1 J 111 #orr 1 - Nodol 35 toyboard Nlth tod ar•s teclUlolaa of unresi-rlct t tvyZloard dnlp lt rt lnly seeas thot it 1bld be posdble CCI dovclop a standoz d koyboard- CRT Q i in 1tlo11 that ceo bo used fot- M ue conecdon for mchln• data - and foT a ctmly readable displ•r with baJTCd cl aracto -• and proper analyst t -oJs 01le- thar i• llot ll opcn•ator cS slped ud approved and b also COf11 1tll1le wlth itandud n• eodlil il Sc Mllq Unfl rtllHtOl ' ID our hOl honcol Orga nil tion thl• has been ioo dll£1cul to do and aa•t IISSIO 101 -'nncx A Scct lon l pa a l 4 states lt carmot lie •Clflh Uhed too stronaly that the suec ou or fall no of 1 ho 4ata rtov od U wl11 depend on tbe operotor trmn crilbeJ' follow• Ing pN4ia 1 y tbc procoduN formu end the cor-reet uso of ta1s os detailed II ere n 11 find who to blAC O not plcktns up the AG· 'l'Z IATS adstake1 in Ulo TEIDUS d ctip The olcl adage •bout ' 1expnleko 1 0t lce9i lftj you q oa ro• paatlng pos a l tws lt only helps you to rocophe t rou 100£ei apin ia cutai C l t n10 in thi cis• h b intno•tlnc to noto bat t bo AD'IA QRS Joint 1111cltaftt atton sroup rdorrod to by tk AG-2Z lATS ls a clatsi euapl o of bus anl tr Ptu l1lp1 1 the SIIUM dOYelopomt coaltt ee - e vs ev toinatlon Obviously very COJSPOtont fenod to b ' Mt Hyland alld t Je toml AG-22 operators •ro rcquirN in or icr for AG• Z2 IATS do -cloia•m vtt• • U shocl by Pa0D 111ltb wor I to l11J11 Uon pa-fectly Collpotoni opotaators can lit Uc tt'° input There seem to • today n0 bo acquird or t n icc Ho111over 1 mdOllletss opcntors MhO Qover mk• • i takes while feeding AG-22 JATS dsra flow h vc not 'Qt been hlro t by Ja'SA plAco loft la PllOD to o • ope-rational dovotopllffltl 111• s ovenetu of thao docis and O m £mMIII en paao UJ Juna Jialy 76 • CRYPIOI OC • l'o o 7 OilfillillHM lllllld I t 1115£ I I I I I f I Non - Responsive J no-July 76 • CRYPTOl 00 age ll UNSI t Glfiilfi _ I Non - Responsive I i- Juno• Tuly 76 • CRYl'TOI QG • P• to 19 UIIClASS IF IED 1 Non - Responsive e911flll lfFl t Juno-Jul ' 76 • CRYPTOLOG • Pta o 20 f Pff5llfillJla L I EO 3 3b 3 PL 86-36 50 USC 3605 ··- THE OOODK 1N··· COLLECTION SYSTEM 'f 1 1 I I I 1nm duat-lan • ' 11ICI GOODEii ColJoction S tti '211 8 hlgbly autODrJ and · fully lntosrotoel •1•u11 or ae qulsltlon lcliuttificat ion ab ion Dlllll t •• I loction Hin • • • · •· • II t conao10·1· c1uu rally 111111anned •re naHsblo for supon -s and traffic 1UU11lyst ed tar The 111' 911 1 1 re bMked tqatroltiy ·- t wo of tbOH san n1 PSPs an ouc oranical ly control led by tlte GOODKIN soft Oro A11 coapcmoats of tllct GDOOUtt $11tca 4 tO Untcd by D data•ttffllll C02 tJ Olltt t 1 ich pcralt interaction UOI I •U CQapancnu on a ti•o•shnln1 bHls I '• t t a nust crSQ n t es a1 a • Uve ·envlron1111nt •1•iast oddltiom L proble-as p Jor to r- rtlifr dc lo 'IDCnts ' 1 '111ic111 '• ' ' i June July 76 • CRYPTOLOC • Page 21 liifiRi11iP81€ - • - • 4 t • _ _ _ • - - I ·-- --------- ----------- l CR F 9f1Nf June-July 7Ci • CllYPl'OLO • Pa ge 21 l lillB 8P81€ EO 3 3b 3 PL 86-36 50 USC 3605 EO 3 3b 3 lilllff IFIII PL 86-36 S0 use 3605 • - - • • •1t I Add-l OJt12 1 f tml ro o Sevoral addlt lo11111 £oaturos of the aJOOllN 1y1tcr1 mko it especial ly attractive to opera- I One o·t the n st a UTa ctive tures for tho Gpcnator is tlu clnastio redueti on i n fi la HI'• vico lrnent Whm a cue ls transferred to a co operator by an a ui ltlon oporotor t he rorur open his her f ile stapl y by depress• Tito COODKUI 1ystca'• softwii TO auto• na tic illy eAttrs th fi c stanc ard-foraat SF t ii a co opon the rue on4 the tollvwin1 Unu ••J c1 L by the operator rtpreseot actual tOP '·or tut tanUvo opontor CGm CIIO Filo cl 6sing ia olso accoapllshed by depTesslna • bfthe opcruor also hlG tho commienco of un O 1 nCI r11e NC l'IY work In1 on an uneluuerecl POsitlop 'Ille GOoollN wort enviroMtllt Ls a paperf're• env ronaent All copy h tben t n ferro4 to dlac Hatj W do produced for naiLon an alysu howevel' fl'OCII the Uno pl'lntor In addltion 1 no tech data 01' toskl clato • niconl- J 1RctUlfeG UH p iltiOII sntee CHH o n nut on lino 4nd l s tmediatoly acco 1lblo to tho opcnu r 11111 11te an AT 'S post ion llttei-o t OfY I a o to msnetle tape eAch Lia e t e operator bl $ rho carriaseqeturft key the GOOOXIN copy J ' tors station ana11st s Gnd rc i icnt of the i ntercept at tlSA i na II koy not trustorr t o 4iisc until the OFer tor hits the file toy •· utuall - 11 t tho end of a June-July 76 • CJYP1'0LOG • PaJO ll SECRiT £Nie ---· -· --------------- • tr - • • • I • • • - · _ _ _ _ _ _ _ _ _ _ _ _ _ _£ _ __ _ _ _ _ _ _ _ _ _ _ _ _ __ IHIIH 811811 Juno•July 76 • CllYnOLOC • Pago 24 ECRif IPIII I EO 3 3b 3 PL 86-36 50 USC 3 60 5 I EO 3 3b 3 PL 86-36 50 USC 3605 8 t R ' OPOtl I I •' • 1be detipert of GDm l JN took a systcsts ap- prouc h to field stttio C'6Uoc1ion All maJol functions a isltl oni i dcntiflcatlon l is1lon IDIWJe Dent collcctton and proce slng • ulCIClaUd to t bo e t t tht ablie-of-the-an •Uow• The GlllCIDJtHf tell' data- i tl'etl COD- ttol lor tt viit al to OOOt JN opera s u ni t c Ystem• W1N 10ftvaN can cc CJ111Plhh theontuembBlon To do'°• opcratin- ac c cpunc e of GOODICTN has uiadllt Dospho 4 are ally i l'ea1cd Wl'kload on the opentor m fotlruo problou htva been onC1 1Qllered Opcrmtor entbuslaS11 •s su•• '1 lllned evtn OlOllj ttlose 111tiD sat at GOODllH JIO• sttlons £01' H lon o a nths tan ' of t ho features that uko the Q ODXIN spcaa espociallt attnctlv t o- opcr' aton vel'e in C Tpoftt ed l nto tho syato H a r Ult of su cat I cm sollc h l d ll and ol t dntd frml srcan smrn -- -· t nr •• req1111ted t'mr Con Ji' ' eaaJ oen ftOa Production to ltCP hit• taken then too f l' frc 11 op11ntioul pro'blt IIS • lioV the ' worry about architecture 'progron mnago• IM tlt 0 'sy tt'lls acqubi tion unate ient and 'stCII pport unasca nt It No ono b lati to do aey real OFtratlanal techl'llqu•s-develo oont 11 0rlc on • J 1119 probJe n -- #t least on • c11lo th t can drlve tl e syst cm-dnotopmftt cyclu our horlicntal orpnlz1tlon l st rUcturo JOCD S to c ml in a CICC h•22 whcno er 10meone Cts good ld and tries to par s11 o it tnrnr z s i CAif SPSlfi I Non - Responsive I UJIGLA5£161 B - June-July 76 4-c'RYPTOLOO • Pago 26 lill612'i lfilii -- 1 Non - Responsive I UIIILAIOIPICB I Non - Responsive runo-July 76 • CRYPICl O j • Pqa 2 ltHlilli l5l 5D -- I HCtMSIPl B IAISbA lilfll I l Non - Responsive I I Non - IOHfilll N l t ' Ju •-July 76 • CRYPTOLOG • Paao 29 r 68UFI iliflt l L Responsive I ••• •• I Non - Responsive r ua uocru•• 1• srsJJ1H1 eu1• ••• ru1 r 111•1111L IOP IIG• A t