Informations de base
Intermsofsourcecoding, Shannonprovedin1951thatwhenthesourceoutputhasredundantmessages, theoutputofthesourcecanbechangedbycodingtomaketheinformationtransmissionrateclosetoChannelcapacity.In1948, ShannonproposedaShannoncodethatcouldmatchthesourceandchannel.In1949, RMFenooftheMassachusettsInstituteofTechnologyproposedFenocoding.In1951, AmericantelecommunicationsengineerD.A.HoffmanproposedamoreeffectiveHuffmancode.Afterthat, facsimilecoding, imagecoding, andvoicecodingappeared.In-depthresearchondatacompressionhassolvedmanypracticalproblemsraisedindigitalcommunications.
Intermsoferrorcorrectioncoding, Shannonproposedanerrorcorrectioncodein1948 (= longueur de code 7, numberofinformationsymbols = 4) .In1949, athree-correctingGraycode erreur sur les bits (longueur de code = 23, numberofinformationsymbols = 12) appeared.In1950, AmericanmathematicianR.W.Hammingpublishedapaper "ErrorDetectionCodesandErrorCorrectionCodes", andproposedthefamousHammingcode, whichhadanimportantinfluenceonerrorcorrectioncodes.Convolutionalcodesappearedin1955.Convolutionalcodesarestillwidelyused. en 1957, cycliccodeswereintroduced.Thestructureofthecycliccodeissimple, itisconvenienttodesignwithalgebratheory, anditisalsoeasytorealize.In1959, HagbergcodesandFaircodesthatcancorrectsuddenerrorsappeared.In1959, RCBossandD.K.RayGiordalioftheUnitedStatesandA.OkungaonofFranceindependentlypublishedafamouscycliccodealmostatthesametime, whichwaslatercalledtheBCHcode (ieBose-Chaudhuri-Hocquenghemcode) .Sequentialdecodingwasproposedin1965.Sequentialdecodinghasbeenusedinspacecommunication.In1967, AJViterbiproposedmaximumlikelihoodconvolutionaldeco ding, whichiscalledViterbidecoding.In1978, vectorcodingappeared.Thevectorcodingmethodisahighlyefficientcodingtechnique.In1980, thenumbertheorymethodwasusedtorealizeReed-Solomoncode (Reed-Solomoncode), referredtoasRScode.Itisactuallyamulti-aryBCHcode.Thiserrorcorrectioncodingtechnologycanreducethenumberofcomponentsoftheencoderintegratedcircuitbyanorderofmagnitude.Ithasbeenwidelyusedinsatellitecommunications.TheconcatenatedcodeconstructedbycombiningRScodeandconvolutionalcodecanbeusedindeepspacecommunication.
Introduction détaillée
Intermsofcryptography, Shannonpublished "CommunicationTheoryofSecrecySystems" en 1949., whichisgenerallyregardedasapioneeringworkincryptography.In1976, DiffiandHermanproposedthepublickeysystemforthefirsttime, whichopenedupanewdirectionfortheresearchofcryptography.Theapplicationofverylargescaleintegratedcircuitsandhigh-speedcomputerspromotesthedevelopmentofsecurecodingtheory, andatthesametimebringsgreatthreatstothesecurityofsecurecommunications.Sincetheintroductionofcomputercomplexitytheoryintocryptographyinthe1970s, soi-calledP-type NP-typeandNP-completeproblemshaveappeared.Thecomplexityfunctionofthealgorithmgrowsexponentially, sothekeyspaceexpands, whichmakestheanalysisandsearchofthecipherfaceseriouschallenges.Cryptographybegantodevelopindepth.
Codage sourceLe codage source généralisé comprend deux aspects : la conversion de l'analogique au numérique (c'est-à-dire la conversion de l'analogique en quantités numériques binaires) et la compression des données (c'est-à-dire l'encodage de ces quantités numériques pour réduire le débit numérique).
Matchingcode
Cette méthode est basée sur la probabilité d'occurrence (distribution de probabilité) de l'objet d'encodage et des codes de longueurs différentes sont donnés, plus la probabilité d'occurrence est élevée, plus la longueur du code donnée est courte.
Codage de transformation
Thismethodistotransformthesignalfirst, fromonesignalspacetoanothersignalspace, andthenencodethetransformedsignal.Transformcodinghasawiderangeofapplicationsinvoiceandimagecoding.Atpresent, thecommonlyusedtransformcodinghastwotypes: predictivecodingandfunctioncoding.Predictivecodingistopredicttheupcomingchangesofthesignalbasedonsomeknownconditionsofthesignal.Itdoesnottransmitthesampledvalueofthesignal, butthedifferencebetweenthesampledvalueofthesignalandthepredictedvalue.Predictivecodingisusedindigitaltelephonesanddigitaltelevisions.ThemostcommonlyusedfunctiontransformationsareFastFourierTransform (FFT), CosineTransform, WalshTransform, HaarTransformandHadamardTransform.Thefrequencyspectrumcharacteristicsofthesignalcanbeobtainedthroughthetransformation, sothedigitalcanbecompressedaccordingtothefrequencyspectrumcharacteristics.
Encodage vectoriel
Cette méthode permet de classer les éventuels messages transmis et de les stocker dans la base de données informatique de l'adresse de réception et l'expéditeur n'envoie que le coiffeur de la base de données pour connaître le contenu du message, ce qui compresse fortement les données envoyées.
Code d'identification
Thismethodismainlyusedforthecodingoftext, symbolsanddatawithstandardshapes.Butvoicecanalsobeidentifiedandcoded.Theroleofidentificationcodeisnotlimitedtocompresseddata, italsohasawiderangeofapplicationsinpatternrecognition.Commonlyusedidentificationmethodsincludeassociationidentificationandlogicalidentification.Identificationcodescangreatlycompressdata.Forexample, usingthevoicerecognitionmethodtotransmitvoice, theaveragebitrateislessthan100bitspersecond.ThemethodofΔmodulatingvoiceisusedtotransmitvoice, andthedigitalratereaches38,400bitspersecond.Thedifferencebetweenthetwoisabout400times.However, theidentificationcoderestoresastandardvoicebasedonacodewhenitisrestored, andcanonlybeusedforspecialtelephonesandquestion-et answerdevicesthatdonotneedtoknowwhothecalleris.Whentheidentificationcodeisusedfortexttransmission, allprintedsymbolsarerecoveredandcanonlybeusedforordinarytelegrams.