用新的高级加密标准AES保持你的数据安全毕业论文外文翻译.docx

用新的高级加密标准AES保持你的数据安全毕业论文外文翻译.docx

《用新的高级加密标准AES保持你的数据安全毕业论文外文翻译.docx》由会员分享，可在线阅读，更多相关《用新的高级加密标准AES保持你的数据安全毕业论文外文翻译.docx（18页珍藏版）》请在冰点文库上搜索。

用新的高级加密标准AES保持你的数据安全毕业论文外文翻译

KeepYourDataSecurewiththeNewAdvancedEncryptionStandard

JamesMcCaffrey

SUMMARY

TheAdvancedEncryptionStandard（AES）isaNationalInstituteofStandardsandTechnologyspecificationfortheencryptionofelectronicdata.Itisexpectedtobecometheacceptedmeansofencryptingdigitalinformation,includingfinancial,telecommunications,andgovernmentdata.ThisarticlepresentsanoverviewofAESandexplainsthealgorithmsituses..AfterreadingthisarticleyouwillbeabletoencryptdatausingAES,testAES-basedsoftware,anduseAESencryptioninyoursystems.

NotethatthecodepresentedinthisarticleandanyotherimplementationbasedonthisarticleissubjecttoapplicableFederalcryptographicmoduleexportcontrols（seeCommercialEncryptionExportControlsfortheexactregulations）.

AESisanewcryptographicalgorithmthatcanbeusedtoprotectelectronicdata.Specifically,AESisaniterative,symmetric-keyblockcipherthatcanusekeysof128,192,and256bits,andencryptsanddecryptsdatainblocksof128bits（16bytes）.Unlikepublic-keyciphers,whichuseapairofkeys,symmetric-keyciphersusethesamekeytoencryptanddecryptdata.Encrypteddatareturnedbyblockciphershavethesamenumberofbitsthattheinputdatahad.Iterativeciphersusealoopstructurethatrepeatedlyperformspermutationsandsubstitutionsoftheinputdata.Figure1showsAESinactionencryptingandthendecryptinga16-byteblockofdatausinga

192-bitkey.

Figure1SomeData

AESisthesuccessortotheolderDataEncryptionStandard（DES）.DESwasapprovedasaFederalstandardin1977andremainedviableuntil1998whenacombinationofadvancesinhardware,software,andcryptanalysistheoryallowedaDES-encryptedmessagetobedecryptedin56hours.SincethattimenumerousothersuccessfulattacksonDES-encrypteddatahavebeenmadeandDESisnowconsideredpastitsusefullifetime.

Inlate1999,theRijndael（pronounced"raindoll"）algorithm,createdbyresearchersJoanDaemenandVincentRijmen,wasselectedbytheNISTastheproposalthatbestmetthedesigncriteriaofsecurity,implementationefficiency,versatility,andsimplicity.AlthoughthetermsAESandRijndaelaresometimesusedinterchangeably,theyaredistinct.AESiswidelyexpectedtobecomethedefactostandardforencryptingallformsofelectronicdataincludingdatausedincommercialapplicationssuchasbankingandfinancialtransactions,telecommunications,andprivateandFederalinformation.

OverviewoftheAESAlgorithm

TheAESalgorithmisbasedonpermutationsandsubstitutions.Permutationsarerearrangementsofdata,andsubstitutionsreplaceoneunitofdatawithanother.AESperformspermutationsandsubstitutionsusingseveraldifferenttechniques.Toillustratethesetechniques,let'swalkthroughaconcreteexampleofAESencryptionusingthedatashowninFigure1.

Thefollowingisthe128-bitvaluethatyouwillencryptwiththeindexesarray:

00112233445566778899aabbccddeeff

0123456789101112131415

The192-bitkeyvalueis:

000102030405060708090a0b0c0d0e0f1011121314151617

01234567891011121314151617181920212223

Figure2Sbox

WhentheAESconstructoriscalled,twotablesthatwillbeusedbytheencryptionmethodareinitialized.ThefirsttableisasubstitutionboxnamedSbox.Itisa16×16matrix.ThefirstfiverowsandcolumnsofSboxareshowninFigure2.Behindthescenes,theencryptionroutinetakesthekeyarrayandusesittogeneratea"keyschedule"tablenamedw[],showninFigure3.

Figure3KeySched.

ThefirstNk（6）rowsofw[]areseededwiththeoriginalkeyvalue（0x00through0x17）andtheremainingrowsaregeneratedfromtheseedkey.ThevariableNkrepresentsthesizeoftheseedkeyin32-bitwords.You'llseeexactlyhoww[]isgeneratedlaterwhenIexaminetheAESimplementation.Thepointisthattherearenowmanykeystouseinsteadofjustone.Thesenewkeysarecalledtheroundkeystodistinguishthemfromtheoriginalseedkey.

Figure4State

TheAESencryptionroutinebeginsbycopyingthe16-byteinputarrayintoa4×4bytematrixnamedState（seeFigure4）.TheAESencryptionalgorithmisnamedCipherandoperatesonState[]andcanbedescribedinpseudocode（seeFigure5）.

Theencryptionalgorithmperformsapreliminaryprocessingstepthat'scalledAddRoundKeyinthespecification.AddRoundKeyperformsabyte-by-byteXORoperationontheStatematrixusingthefirstfourrowsofthekeyschedule,andXORsinputState[r,c]withroundkeystablew[c,r].

Forexample,ifthefirstrowoftheStatematrixholdsthebytes{00,44,88,cc},andthefirstcolumnofthekeyscheduleis{00,04,08,0c},thenthenewvalueofState[0,2]istheresultofXORingState[0,2]（0x88）withw[2,0]（0x08）,or0x80:

10001000

00001000XOR

10000000

ThemainloopoftheAESencryptionalgorithmperformsfourdifferentoperationsontheStatematrix,calledSubBytes,ShiftRows,MixColumns,andAddRoundKeyinthespecification.TheAddRoundKeyoperationisthesameasthepreliminaryAddRoundKeyexceptthateachtimeAddRoundKeyiscalled,thenextfourrowsofthekeyscheduleareused.TheSubBytesroutineisasubstitutionoperationthattakeseachbyteintheStatematrixandsubstitutesanewbytedeterminedbytheSboxtable.Forexample,ifthevalueofState[0,1]is0x40andyouwanttofinditssubstitute,youtakethevalueatState[0,1]（0x40）andletxequaltheleftdigit（4）andyequaltherightdigit（0）.ThenyouusexandyasindexesintotheSboxtabletofindthesubstitutionvalue,asshowninFigure2.

ShiftRowsisapermutationoperationthatrotatesbytesintheStatematrixtotheleft.Figure6showshowShiftRowsworksonState[].Row0ofStateisrotated0positionstotheleft,row1isrotated1positionleft,row2isrotated2positionsleft,androw3isrotated3positionsleft.

Figure6RunningShiftRowsonState

TheMixColumnsoperationisasubstitutionoperationthatisthetrickiestpartoftheAESalgorithmtounderstand.Itreplaceseachbytewiththeresultofmathematicalfieldadditionsandmultiplicationsofvaluesinthebyte'scolumn.Iwillexplainthedetailsofspecialfieldadditionandmultiplicationinthenextsection.

SupposethevalueatState[0,1]is0x09,andtheothervaluesincolumn1are0x60,0xe1,and0x04;thenthenewvalueforState[0,1]isshowninthefollowing:

State[0,1]=（State[0,1]*0x01）+（State[1,1]*0x02）+（State[2,1]*0x03）+（State[3,1]*0x01）=（0x09*0x01）+（0x60*0x02）+（0xe1*0x03）+（0x04*0x01）

=0x57

Theadditionandmultiplicationarespecialmathematicalfieldoperations,nottheusualadditionandmultiplicationonintegers.

ThefouroperationsSubBytes,ShiftRows,MixColumns,andAddRoundKeyarecalledinsidealoopthatexecutesNrtimes—thenumberofroundsforagivenkeysize,less1.Thenumberofroundsthattheencryptionalgorithmusesiseither10,12,or14anddependsonwhethertheseedkeysizeis128,192,or256bits.Inthisexample,becauseNrequals12,thefouroperationsarecalled11times.Afterthisiterationcompletes,theencryptionalgorithmfinishesbycallingSubBytes,ShiftRows,andAddRoundKeybeforecopyingtheStatematrixtotheoutputparameter.

Insummary,therearefouroperationsthatareattheheartoftheAESencryptionalgorithm.AddRoundKeysubstitutesgroupsof4bytesusingroundkeysgeneratedfromtheseedkeyvalue.SubBytessubstitutesindividualbytesusingasubstitutiontable.ShiftRowspermutesgroupsof4bytesbyrotating4-byterows.MixColumnssubstitutesbytesusingacombinationofbothfieldadditionandmultiplication.

FieldAdditionandMultiplicationinGF（28）

Asyou'veseen,theAESencryptionalgorithmusesfairlystraightforwardtechniquesforsubstitutionandpermutation,exceptfortheMixColumnsroutine.TheMixColumnsroutineusesspecialadditionandmultiplication.TheadditionandmultiplicationusedbyAESarebasedonmathematicalfieldtheory.Inparticular,AESisbasedonafieldcalledGF（28）.

TheGF（28）fieldconsistsofasetof256valuesfrom0x00to0xff,plusadditionandmultiplication,hencethe（28）.GFstandsforGaloisField,namedafterthemathematicianwhofoundedfieldtheory.OneofthecharacteristicsofGF（28）isthattheresultofanadditionormultiplicationoperationmustbeintheset{0x00...0xff}.Althoughthetheoryoffieldsisratherdeep,thenetresultforGF（28）additionissimple:

GF（28）additionisjusttheXORoperation.

MultiplicationinGF（28）istrickier,however.Asyou'llseelaterintheC#implementation,theAESencryptionanddecryptionroutinesneedtoknowhowtomultiplybyonlythesevenconstants0x01,0x02,0x03,0x09,0x0b,0x0d,and0x0e.SoinsteadofexplainingGF（28）multiplicationtheoryingeneral,Iwillexplainitjustforthesesevenspecificcases.

Multiplicationby0x01inGF（28）isspecial;itcorrespondstomultiplicationby1innormalarithmeticandworksthesameway—anyvaluetimes0x01equalsitself.

Nowlet'slookatmultiplicationby0x02.Asinthecaseofaddition,thetheoryisdeep,butthenetresultisfairlysimple.Ifthevaluebeingmultipliedislessthan0x80,thentheresultofmultiplicationisjustthevalueleft-shifted1bitposition.Ifthevaluebeingmultipliedisgreaterthanorequalto0x80,thentheresultofmultiplicationisthevalueleft-shifted1bitpositionXORedwiththevalue0x1b.Thisprevents"fieldoverflow"andkeepstheproductofthemultiplicationinrange.

Onceyou'veestablishedadditionandmultiplicationby0x02inGF（28）,youcanusethemtodefinemultiplicationbyanyconstant.Tomult