2017 암호경진대회 1번

알파벳으로 구성된 평문을 2개의 값으로 shift하는 key table로 암호화한 암호문이 주어졌을때 평문을 찾는 문제입니다.

소문제 1번은 경우 암호문길이가 길어서 풀 수 있었는데, 2번의 경우 길이가 너무 짧아 못풀었습니다.

제출한 답안 그대로 올릴겠습니다

1-1.

복구된 평문:

youneedtomanageyourresourcesspeednolongermattersitusedtobethecasethatprogramstookareallylongtimetoruncpuswereexpensivememorywasexpensiverunningtimeofaprogramusedtobeanimportantmetriccomputerswereveryexpensiveandsowastheelectricitytorunthemoptimizationoftheseresourceswasdonebecauseofaneternalbusinesslawoptimizeyourmostexpensiveresourcehistoricallythemostexpensiveresourcewascomputerruntimethisiswhatleadtothestudyofcomputersciencewhichfocusesonefficiencyofdifferentalgorithmshoweverthisisnolongertrueassiliconisnowcheaplikereallycheapruntimeisnolongeryourmostexpensiveresourceacompanysmostexpensiveresourceisnowitsemployeestimeorinotherwordsyouitsmoreimportanttogetstuffdonethantomakeitgofastinfactthisissoimportantiamgoingtoputitagainrighthereasifitwasaquoteforthosewhoarejustbrowsingitsmoreimportanttogetstuffdonethantomakeitgofastyoumightbesayingmycompanycaresaboutspeedibuildawebapplicationandallresponseshavetobefasterthanxmillisecondsorwehavehadcustomerscancelbecausetheythinkourappistooslowiamnottryingtosaythatspeeddoesntmatteratalliamsimplytryingtosaythatitsnolongerthemostimportantthingitsnotyourmostexpensiveresourcespeedspeedistheonlythingthatmatterswhenyousayspeedinthecontextofprogrammingyoutypicallymeanperformanceakacpucycleswhenyourceosaysspeedinthecontextofprogramminghemeansbusinessspeedthemostimportantmetricistimetomarketultimatelyitdoesntmatterhowfastyourproductwebappisitdoesntmatterwhatlanguageitswritteninitdoesntevenmatterhowmuchmoneyittakestorunattheendofthedaytheonethingthatwillmakeyourcompanysurviveordieistimetomarketimnotjusttalkingaboutthestartupideaofhowlongittakestillyoumakemoneybutmoresothetimeframeoffromideatocustomershandstheonlywaytosurviveinbusinessistoinnovatefasterthanyourcompetitorsitdoesntmatterhowmanygoodideasyoucomeupwithifyourcompetitorsshipbeforeyoudoyouhavetobethefirsttomarketoratleastkeepuponceyouslowdownyouaredonetheonlywaytosurviveinbusinessistoinnovatefasterthanyourcompetitorsacaseofmicroservicescompanieslikeamazongoogleandnetflixunderstandtheimportanceofmovingfasttheyhavecreatedabusinesssystemwheretheycanmovefastandinnovatequicklymicroservicesarethesolutiontotheirproblemthisarticlehasnothingtodowithwhetherornotyoushouldbeusingmicroservicesbutatleastacceptthatamazonandgooglethinktheyshouldbeusingthemmicroservicesareinherentlyslowtheveryconceptofamicroserviceistobreakupaboundarybyanetworkcallthismeansyouaretakingwhatwasafunctioncallacouplecpucyclesandturningitintoanetworkcallthereisntmuchyoucoulddothatisworseintermsofperformancenetworkcallsarereallyslowcomparedtothecpubutthesebigcompaniesstillchoosetousemicroservicestherereallyisntanarchitectureslowerthanmicroservicesthatiknowofmicroservicesbiggestconisperformancebutgreatestproistimetomarketbybuildingteamsaroundsmallerprojectsandcodebasesacompanyisabletoiterateandinnovateatamuchfasterpacethisjustgoestoshowthatverylargecompaniesalsocareabouttimetomarketnotjuststartupscpuisnotyourbottleneckifyouwriteanetworkapplicationsuchasawebserverchancesarecputimeisnotthebottleneckofyourapplicationwhenyourwebserverhandlesarequestitprobablymakesacouplenetworkcallssuchastoyourdatabaseorperhapsacacheserverlikerediswhiletheseservicesthemselvesmaybefastthenetworkcalltothemisslowthereisareallygreatblogarticleonthespeeddifferencesofcertainoperationsinthearticletheauthorscalescpucycletimesintomoreunderstandablehumantimesifasinglecpucyclewastheequivalentofonesecondthenanetworkcallfromcaliforniatonewyorkwouldbetheequivalentoffouryearsthatishowmuchslowernetworkisforsomeroughestimatesletssayanormalnetworkcallinsidethesamedatacentertakesaboutthreemsthatwouldbetheequivalentofthreemonthsinourhumanscalenowimagineyourprogramisverycpuintensiveittakesahundredthousandcyclestorespondtoasinglecallthatwouldbetheequivalentofjustoveronedaynowletssayyouusealanguagethatisfivetimesasslownowittakesaboutfivedayswellcomparethattoourthreemonthnetworkcallandthefourdaydifferencedoesntreallymattermuchatallifsomeonehastowaitatleastthreemonthsforapackageidontthinkanextrafourdayswillreallymatterallthatmuchtothemwhatthisultimatelymeansisthatevenifpythonisslowitdoesntmatterthespeedofthelanguageorcputimeisalmostnevertheissuegoogleactuallydidastudyonthisveryconceptandtheywroteapaperonitthepapertalksaboutdesigningahighthroughputsystemintheconclusiontheysayitmayseemparadoxicaltou

테이블

A

B

C

D

E

F

G

H

I

J

K

L

M

N

O

P

Q

R

S

T

U

V

W

X

Y

Z

v

f

c

p

k

g

q

n

h

e

u

y

z

s

o

a

w

b

d

r

x

m

i

t

j

l

시프트 값 : 13, 21

풀이:

테이블이 n번 움직인 후에는 초기의 테이블과 같은 모양을 띌 것으로 생각했다. 그리고 길이 n위치마다 있는 알파벳은 같은 키테이블로 암호화되기 때문에 일반적인 알파벳의 빈도를 따를 것으로 생각했다. 

반복되는 길이를 추측하기 위해 암호문을 n의 길이로 나누고 같은 위치를 같은 문자끼리 합하였다. 

그리고 index of coincidence를 계산하였다. 그 결과 키 길이(테이블이 처음과 같은 상태로 돌아오는 길이)에서 일반적인 알파벳의 index of coincidence값인 0.6정도가 나왔다. 

이후 암호문을 26의 길이잡고 알파벳의 빈도를 분석하였다. 빈도 분석 후에 자주 사용되는 2글자, 3글자를 대입해봄으로써 비밀정보를 구할 수 있었다.

1-2.

답안:

문제 1-1과 같은 방법으로 접근하여 반복되는 길이가 26인 것은 확인하였으나. 그 외의 비밀정보는 찾지 못하였다. 1-1에 비해 주어진 데이터가 적어 구하기 못하였다.