比特幣行情 比特幣行情
Ctrl+D 比特幣行情
ads

atc:比特幣白皮書(英文版)

Author:

Time:1900/1/1 0:00:00

比特幣白皮書

Bitcoin:APeer-to-PeerElectronicCashSystem

SatoshiNakamoto

satoshin@gmx.com

www.bitcoin.org

Abstract

Bitcoin:APeer-to-PeerElectronicCashSystem

SatoshiNakamoto

satoshin@gmx.com

www.bitcoin.org

1.Introduction

CommerceontheInternethascometorelyalmostexclusivelyonfinancialinstitutionsservingastrustedthirdpartiestoprocesselectronicpayments.Whilethesystemworkswellenoughformosttransactions,itstillsuffersfromtheinherentweaknessesofthetrustbasedmodel.Completelynon-reversibletransactionsarenotreallypossible,sincefinancialinstitutionscannotavoidmediatingdisputes.Thecostofmediationincreasestransactioncosts,limitingtheminimumpracticaltransactionsizeandcuttingoffthepossibilityforsmallcasualtransactions,andthereisabroadercostinthelossofabilitytomakenon-reversiblepaymentsfornonreversibleservices.Withthepossibilityofreversal,theneedfortrustspreads.Merchantsmustbewaryoftheircustomers,hasslingthemformoreinformationthantheywouldotherwiseneed.Acertainpercentageoffraudisacceptedasunavoidable.Thesecostsandpaymentuncertaintiescanbeavoidedinpersonbyusingphysicalcurrency,butnomechanismexiststomakepaymentsoveracommunicationschannelwithoutatrustedparty.

Whatisneededisanelectronicpaymentsystembasedoncryptographicproofinsteadoftrust,allowinganytwowillingpartiestotransactdirectlywitheachotherwithouttheneedforatrustedthirdparty.Transactionsthatarecomputationallyimpracticaltoreversewouldprotectsellersfromfraud,androutineescrowmechanismscouldeasilybeimplementedtoprotectbuyers.Inthispaper,weproposeasolutiontothedouble-spendingproblemusingapeer-to-peerdistributedtimestampservertogeneratecomputationalproofofthechronologicalorderoftransactions.ThesystemissecureaslongashonestnodescollectivelycontrolmoreCPUpowerthananycooperatinggroupofattackernodes.

2.Transactions

Wedefineanelectroniccoinasachainofdigitalsignatures.Eachownertransfersthecointothenextbydigitallysigningahashoftheprevioustransactionandthepublickeyofthenextownerandaddingthesetotheendofthecoin.Apayeecanverifythesignaturestoverifythechainofownership.

比特幣錨定幣WBTC鎖倉量突破100億美元 創歷史新高:金色財經報道,據最新數據顯示,比特幣錨定幣WBTC鎖倉量已突破100億美元,創下歷史新高,本文撰寫時為10,092,458,454美元,發行量為198,623枚。比特幣錨定幣是發行在非比特幣網絡上的價格錨定于原生比特幣的一類特殊代幣。[2021/9/3 22:58:42]

Theproblemofcourseisthepayeecan'tverifythatoneoftheownersdidnotdouble-spendthecoin.Acommonsolutionistointroduceatrustedcentralauthority,ormint,thatcheckseverytransactionfordoublespending.Aftereachtransaction,thecoinmustbereturnedtotheminttoissueanewcoin,andonlycoinsissueddirectlyfromthemintaretrustednottobedouble-spent.Theproblemwiththissolutionisthatthefateoftheentiremoneysystemdependsonthecompanyrunningthemint,witheverytransactionhavingtogothroughthem,justlikeabank.

Weneedawayforthepayeetoknowthatthepreviousownersdidnotsignanyearliertransactions.Forourpurposes,theearliesttransactionistheonethatcounts,sowedon'tcareaboutlaterattemptstodouble-spend.Theonlywaytoconfirmtheabsenceofatransactionistobeawareofalltransactions.Inthemintbasedmodel,themintwasawareofalltransactionsanddecidedwhicharrivedfirst.Toaccomplishthiswithoutatrustedparty,transactionsmustbepubliclyannounced,andweneedasystemforparticipantstoagreeonasinglehistoryoftheorderinwhichtheywerereceived.Thepayeeneedsproofthatatthetimeofeachtransaction,themajorityofnodesagreeditwasthefirstreceived.

3.TimestampServer

Thesolutionweproposebeginswithatimestampserver.Atimestampserverworksbytakingahashofablockofitemstobetimestampedandwidelypublishingthehash,suchasinanewspaperorUsenetpost.Thetimestampprovesthatthedatamusthaveexistedatthetime,obviously,inordertogetintothehash.Eachtimestampincludestheprevioustimestampinitshash,formingachain,witheachadditionaltimestampreinforcingtheonesbeforeit.

4.Proof-of-Work

Toimplementadistributedtimestampserveronapeer-to-peerbasis,wewillneedtouseaproofof-worksystemsimilartoAdamBack'sHashcash,ratherthannewspaperorUsenetposts.Theproof-of-workinvolvesscanningforavaluethatwhenhashed,suchaswithSHA-256,thehashbeginswithanumberofzerobits.Theaverageworkrequiredisexponentialinthenumberofzerobitsrequiredandcanbeverifiedbyexecutingasinglehash.

Forourtimestampnetwork,weimplementtheproof-of-workbyincrementinganonceintheblockuntilavalueisfoundthatgivestheblock'shashtherequiredzerobits.OncetheCPUefforthasbeenexpendedtomakeitsatisfytheproof-of-work,theblockcannotbechangedwithoutredoingthework.Aslaterblocksarechainedafterit,theworktochangetheblockwouldincluderedoingalltheblocksafterit.

比特幣挖礦難度預計在15個小時后下降14.34%:金色財經報道,據BTC.com顯示,目前比特幣全網算力在95.31EH/s,比特幣挖礦難度預計將在15個小時后下調14.34%,由目前的16.55T下降到14.18T左右。上一次挖礦難度出現大幅下降發生在2018年12月,當時比特幣價格位于3000-4000美元位置。[2020/3/25]

Theproof-of-workalsosolvestheproblemofdeterminingrepresentationinmajoritydecisionmaking.Ifthemajoritywerebasedonone-IP-address-one-vote,itcouldbesubvertedbyanyoneabletoallocatemanyIPs.Proof-of-workisessentiallyone-CPU-one-vote.Themajoritydecisionisrepresentedbythelongestchain,whichhasthegreatestproofof-workeffortinvestedinit.IfamajorityofCPUpoweriscontrolledbyhonestnodes,thehonestchainwillgrowthefastestandoutpaceanycompetingchains.Tomodifyapastblock,anattackerwouldhavetoredotheproof-ofworkoftheblockandallblocksafteritandthencatchupwithandsurpasstheworkofthehonestnodes.Wewillshowlaterthattheprobabilityofaslowerattackercatchingupdiminishesexponentiallyassubsequentblocksareadded.

Tocompensateforincreasinghardwarespeedandvaryinginterestinrunningnodesovertime,theproof-of-workdifficultyisdeterminedbyamovingaveragetargetinganaveragenumberofblocksperhour.Ifthey'regeneratedtoofast,thedifficultyincreases.

5.Network

Thestepstorunthenetworkareasfollows:

1)Newtransactionsarebroadcasttoallnodes.

2)Eachnodecollectsnewtransactionsintoablock.

3)Eachnodeworksonfindingadifficultproof-of-workforitsblock.

4)Whenanodefindsaproof-of-work,itbroadcaststheblocktoallnodes.

5)Nodesaccepttheblockonlyifalltransactionsinitarevalidandnotalreadyspent.

6)Nodesexpresstheiracceptanceoftheblockbyworkingoncreatingthenextblockinthechain,usingthehashoftheacceptedblockastheprevioushash.

Nodesalwaysconsiderthelongestchaintobethecorrectoneandwillkeepworkingonextendingit.Iftwonodesbroadcastdifferentversionsofthenextblocksimultaneously,somenodesmayreceiveoneortheotherfirst.Inthatcase,theyworkonthefirstonetheyreceived,butsavetheotherbranchincaseitbecomeslonger.Thetiewillbebrokenwhenthenextproofof-workisfoundandonebranchbecomeslonger;thenodesthatwereworkingontheotherbranchwillthenswitchtothelongerone.

Newtransactionbroadcastsdonotnecessarilyneedtoreachallnodes.Aslongastheyreachmanynodes,theywillgetintoablockbeforelong.Blockbroadcastsarealsotolerantofdroppedmessages.Ifanodedoesnotreceiveablock,itwillrequestitwhenitreceivesthenextblockandrealizesitmissedone.

聲音 | 黃金支持者Peter Schiff:美聯儲創造了比特幣泡沫:當被問及看跌比特幣的原因,黃金支持者、Euro Pacific Capital首席執行官Peter Schiff首先指出“美聯儲創造了比特幣泡沫”,就像他們創造網絡泡沫和房地產泡沫一樣。然而,他表示,這不是完全相同的方式,但肯定這是對美聯儲貨幣政策錯誤的反應。(AMBCrypto)[2019/10/28]

6.Incentive

Byconvention,thefirsttransactioninablockisaspecialtransactionthatstartsanewcoinownedbythecreatoroftheblock.Thisaddsanincentivefornodestosupportthenetwork,andprovidesawaytoinitiallydistributecoinsintocirculation,sincethereisnocentralauthoritytoissuethem.Thesteadyadditionofaconstantofamountofnewcoinsisanalogoustogoldminersexpendingresourcestoaddgoldtocirculation.Inourcase,itisCPUtimeandelectricitythatisexpended.

Theincentivecanalsobefundedwithtransactionfees.Iftheoutputvalueofatransactionislessthanitsinputvalue,thedifferenceisatransactionfeethatisaddedtotheincentivevalueoftheblockcontainingthetransaction.Onceapredeterminednumberofcoinshaveenteredcirculation,theincentivecantransitionentirelytotransactionfeesandbecompletelyinflationfree.

Theincentivemayhelpencouragenodestostayhonest.IfagreedyattackerisabletoassemblemoreCPUpowerthanallthehonestnodes,hewouldhavetochoosebetweenusingittodefraudpeoplebystealingbackhispayments,orusingittogeneratenewcoins.Heoughttofinditmoreprofitabletoplaybytherules,suchrulesthatfavourhimwithmorenewcoinsthaneveryoneelsecombined,thantounderminethesystemandthevalidityofhisownwealth.

7.ReclaimingDiskSpace

Oncethelatesttransactioninacoinisburiedunderenoughblocks,thespenttransactionsbeforeitcanbediscardedtosavediskspace.Tofacilitatethiswithoutbreakingtheblock'shash,transactionsarehashedinaMerkleTree,withonlytherootincludedintheblock'shash.Oldblockscanthenbecompactedbystubbingoffbranchesofthetree.Theinteriorhashesdonotneedtobestored.

Ablockheaderwithnotransactionswouldbeabout80bytes.Ifwesupposeblocksaregeneratedevery10minutes,80bytes*6*24*365=4.2MBperyear.Withcomputersystemstypicallysellingwith2GBofRAMasof2008,andMoore'sLawpredictingcurrentgrowthof1.2GBperyear,storageshouldnotbeaproblemeveniftheblockheadersmustbekeptinmemory.

8.SimplifiedPaymentVerification

Itispossibletoverifypaymentswithoutrunningafullnetworknode.Auseronlyneedstokeepacopyoftheblockheadersofthelongestproof-of-workchain,whichhecangetbyqueryingnetworknodesuntilhe'sconvincedhehasthelongestchain,andobtaintheMerklebranchlinkingthetransactiontotheblockit'stimestampedin.Hecan'tcheckthetransactionforhimself,butbylinkingittoaplaceinthechain,hecanseethatanetworknodehasacceptedit,andblocksaddedafteritfurtherconfirmthenetworkhasacceptedit.

聲音 | 李啟威:巴菲特現在和將來都不會對比特幣感興趣 不可能說服他:據AMBCrypto消息,談到萊特幣最近和即將到來的一些發展,萊特幣創始人李啟威將萊特幣的成功歸因于即將到來的減半。此外,當被問及與沃倫·巴菲特共進午餐的計劃時,李啟威強調,他的目標不是說服巴菲特支持加密貨幣。“不可能說服他(巴菲特)。他不投資科技方面的東西,所以他現在和將來都不會對比特幣感興趣。這一點也不奇怪。”[2019/7/19]

Assuch,theverificationisreliableaslongashonestnodescontrolthenetwork,butismorevulnerableifthenetworkisoverpoweredbyanattacker.Whilenetworknodescanverifytransactionsforthemselves,thesimplifiedmethodcanbefooledbyanattacker'sfabricatedtransactionsforaslongastheattackercancontinuetooverpowerthenetwork.Onestrategytoprotectagainstthiswouldbetoacceptalertsfromnetworknodeswhentheydetectaninvalidblock,promptingtheuser'ssoftwaretodownloadthefullblockandalertedtransactionstoconfirmtheinconsistency.Businessesthatreceivefrequentpaymentswillprobablystillwanttoruntheirownnodesformoreindependentsecurityandquickerverification.

9.CombiningandSplittingValue

Althoughitwouldbepossibletohandlecoinsindividually,itwouldbeunwieldytomakeaseparatetransactionforeverycentinatransfer.Toallowvaluetobesplitandcombined,transactionscontainmultipleinputsandoutputs.Normallytherewillbeeitherasingleinputfromalargerprevioustransactionormultipleinputscombiningsmalleramounts,andatmosttwooutputs:oneforthepayment,andonereturningthechange,ifany,backtothesender.

Itshouldbenotedthatfan-out,whereatransactiondependsonseveraltransactions,andthosetransactionsdependonmanymore,isnotaproblemhere.Thereisnevertheneedtoextractacompletestandalonecopyofatransaction'shistory.

10.Privacy

Thetraditionalbankingmodelachievesalevelofprivacybylimitingaccesstoinformationtothepartiesinvolvedandthetrustedthirdparty.Thenecessitytoannouncealltransactionspubliclyprecludesthismethod,butprivacycanstillbemaintainedbybreakingtheflowofinformationinanotherplace:bykeepingpublickeysanonymous.Thepubliccanseethatsomeoneissendinganamounttosomeoneelse,butwithoutinformationlinkingthetransactiontoanyone.Thisissimilartothelevelofinformationreleasedbystockexchanges,wherethetimeandsizeofindividualtrades,the"tape",ismadepublic,butwithouttellingwhothepartieswere.

Asanadditionalfirewall,anewkeypairshouldbeusedforeachtransactiontokeepthemfrombeinglinkedtoacommonowner.Somelinkingisstillunavoidablewithmulti-inputtransactions,whichnecessarilyrevealthattheirinputswereownedbythesameowner.Theriskisthatiftheownerofakeyisrevealed,linkingcouldrevealothertransactionsthatbelongedtothesameowner.

聲音 | 金融評論家:比特幣在走自己的路 挑戰它的人看起來有些愚蠢:據CCN消息,金融評論家Keizer最近在一檔節目中表示:“比特幣正在走自己的路,它能讓挑戰它的人看起來非常愚蠢。當然在比特幣這個領域里,你有Craig Wright,你有Roger Ver,你有Mike Hearn,其實這些類型的人都患有比特幣紊亂癥。在過去的幾年里,Nouriel Roubini和Paul Krugman等主流經濟學家,看起來更加愚蠢,因為他們從來沒有花時間研究比特幣是什么。”[2019/4/17]

11.Calculations

Weconsiderthescenarioofanattackertryingtogenerateanalternatechainfasterthanthehonestchain.Evenifthisisaccomplished,itdoesnotthrowthesystemopentoarbitrarychanges,suchascreatingvalueoutofthinairortakingmoneythatneverbelongedtotheattacker.Nodesarenotgoingtoacceptaninvalidtransactionaspayment,andhonestnodeswillneveracceptablockcontainingthem.Anattackercanonlytrytochangeoneofhisowntransactionstotakebackmoneyherecentlyspent.

TheracebetweenthehonestchainandanattackerchaincanbecharacterizedasaBinomialRandomWalk.Thesuccesseventisthehonestchainbeingextendedbyoneblock,increasingitsleadby+1,andthefailureeventistheattacker'schainbeingextendedbyoneblock,reducingthegapby-1.

TheprobabilityofanattackercatchingupfromagivendeficitisanalogoustoaGambler'sRuinproblem.Supposeagamblerwithunlimitedcreditstartsatadeficitandplayspotentiallyaninfinitenumberoftrialstotrytoreachbreakeven.Wecancalculatetheprobabilityheeverreachesbreakeven,orthatanattackerevercatchesupwiththehonestchain,asfollows:

Givenourassumptionthatp>q,theprobabilitydropsexponentiallyasthenumberofblockstheattackerhastocatchupwithincreases.Withtheoddsagainsthim,ifhedoesn'tmakealuckylungeforwardearlyon,hischancesbecomevanishinglysmallashefallsfurtherbehind.

Wenowconsiderhowlongtherecipientofanewtransactionneedstowaitbeforebeingsufficientlycertainthesendercan'tchangethetransaction.Weassumethesenderisanattackerwhowantstomaketherecipientbelievehepaidhimforawhile,thenswitchittopaybacktohimselfaftersometimehaspassed.Thereceiverwillbealertedwhenthathappens,butthesenderhopesitwillbetoolate.

Thereceivergeneratesanewkeypairandgivesthepublickeytothesendershortlybeforesigning.Thispreventsthesenderfrompreparingachainofblocksaheadoftimebyworkingonitcontinuouslyuntilheisluckyenoughtogetfarenoughahead,thenexecutingthetransactionatthatmoment.Oncethetransactionissent,thedishonestsenderstartsworkinginsecretonaparallelchaincontaininganalternateversionofhistransaction.

Therecipientwaitsuntilthetransactionhasbeenaddedtoablockandzblockshavebeenlinkedafterit.Hedoesn'tknowtheexactamountofprogresstheattackerhasmade,butassumingthehonestblockstooktheaverageexpectedtimeperblock,theattacker'spotentialprogresswillbeaPoissondistributionwithexpectedvalue:

Togettheprobabilitytheattackercouldstillcatchupnow,wemultiplythePoissondensityforeachamountofprogresshecouldhavemadebytheprobabilityhecouldcatchupfromthatpoint:

Rearrangingtoavoidsummingtheinfinitetailofthedistribution...

ConvertingtoCcode...

#includedoubleAttackerSuccessProbability(doubleq,intz)

{

doublep=1.0-q;

doublelambda=z*(q/p);

doublesum=1.0;

inti,k;

for(k=0;k<=z;k++)

{

doublepoisson=exp(-lambda);

for(i=1;i<=k;i++)

poisson*=lambda/i;

sum-=poisson*(1-pow(q/p,z-k));

}

returnsum;

}

Runningsomeresults,wecanseetheprobabilitydropoffexponentiallywithz.

q=0.1

z=0P=1.0000000

z=1P=0.2045873

z=2P=0.0509779

z=3P=0.0131722

z=4P=0.0034552

z=5P=0.0009137

z=6P=0.0002428

z=7P=0.0000647

z=8P=0.0000173

z=9P=0.0000046

z=10P=0.0000012

q=0.3

z=0P=1.0000000

z=5P=0.1773523

z=10P=0.0416605

z=15P=0.0101008

z=20P=0.0024804

z=25P=0.0006132

z=30P=0.0001522

z=35P=0.0000379

z=40P=0.0000095

z=45P=0.0000024

z=50P=0.0000006

SolvingforPlessthan0.1%...

P<0.001

q=0.10z=5

q=0.15z=8

q=0.20z=11

q=0.25z=15

q=0.30z=24

q=0.35z=41

q=0.40z=89

q=0.45z=340

12.Conclusion

Wehaveproposedasystemforelectronictransactionswithoutrelyingontrust.Westartedwiththeusualframeworkofcoinsmadefromdigitalsignatures,whichprovidesstrongcontrolofownership,butisincompletewithoutawaytopreventdouble-spending.Tosolvethis,weproposedapeer-to-peernetworkusingproof-of-worktorecordapublichistoryoftransactionsthatquicklybecomescomputationallyimpracticalforanattackertochangeifhonestnodescontrolamajorityofCPUpower.Thenetworkisrobustinitsunstructuredsimplicity.Nodesworkallatoncewithlittlecoordination.Theydonotneedtobeidentified,sincemessagesarenotroutedtoanyparticularplaceandonlyneedtobedeliveredonabesteffortbasis.Nodescanleaveandrejointhenetworkatwill,acceptingtheproof-ofworkchainasproofofwhathappenedwhiletheyweregone.TheyvotewiththeirCPUpower,expressingtheiracceptanceofvalidblocksbyworkingonextendingthemandrejectinginvalidblocksbyrefusingtoworkonthem.Anyneededrulesandincentivescanbeenforcedwiththisconsensusmechanism.

References

W.Dai,"b-money,"http://www.weidai.com/bmoney.txt,1998.

H.Massias,X.S.Avila,andJ.-J.Quisquater,"Designofasecuretimestampingservicewithminimal

trustrequirements,"In20thSymposiumonInformationTheoryintheBenelux,May1999.

S.Haber,W.S.Stornetta,"Howtotime-stampadigitaldocument,"InJournalofCryptology,vol3,no

2,pages99-111,1991.

D.Bayer,S.Haber,W.S.Stornetta,"Improvingtheefficiencyandreliabilityofdigitaltime-stamping,"

InSequencesII:MethodsinCommunication,SecurityandComputerScience,pages329-334,1993.

S.Haber,W.S.Stornetta,"Securenamesforbit-strings,"InProceedingsofthe4thACMConference

onComputerandCommunicationsSecurity,pages28-35,April1997.

A.Back,"Hashcash-adenialofservicecounter-measure,"

http://www.hashcash.org/papers/hashcash.pdf,2002.

R.C.Merkle,"Protocolsforpublickeycryptosystems,"InProc.1980SymposiumonSecurityand

Privacy,IEEEComputerSociety,pages122-133,April1980.

W.Feller,"Anintroductiontoprobabilitytheoryanditsapplications,"1957.

沙棘財經是沙棘傳媒旗下專注大數據、人工智能、區塊鏈、幣圈的深度報道的垂直自媒體。微信公眾號:shaji-media

Tags:atcCPUtronatc幣是什么幣CPU幣CPU價格tron幣是什么幣

歐易交易所app下載
區塊鏈:密碼學,支持區塊鏈和比特幣的上層建筑

如果要預測2018年最熱門的事件,區塊鏈和比特幣無疑可以算得上是其中之一,就連比特幣的發明人——神秘地消失在互聯網世界的中本聰,甚至都成為2018年《時代》周刊年度人物的有力競爭者.

1900/1/1 0:00:00
比特幣:區塊鏈大火如何應對:看我與比特幣擦肩參透的貪婪與恐懼

區塊鏈依舊火熱,仿佛黃河之水天上來,奔流不停息。比特幣像在坐過山車,忽上忽下起伏,陰晴不定。機構們爭先恐后要搭乘區塊鏈的時代列車,Allin一往無前.

1900/1/1 0:00:00
ENS:北汽BJ90到店,配4.0T引擎+柏林之聲,售百萬,不愧最貴國產SUV

國產車一直以來給人的感覺基本都是中低端車型居多,除了售價高達500萬的紅旗L5,國產車在高端市場基本沒有拿得出手的車型。特別是在豪華SUV上,整個市場已經全部被國外品牌給占領.

1900/1/1 0:00:00
比特幣:幣圈進化,IFO、IHO玩法更高級,李笑來Candy“糖果”野望未來

如果說“鏈圈一日,人間一月”是區塊鏈創業者們的感嘆,那么“幣圈一日,人間一年”則是玩幣一族對數字貨幣的驚嘆。不管是區塊鏈創業者還是數字貨幣玩家,如果不緊跟節奏,很快就會成為落后分子.

1900/1/1 0:00:00
新經幣:價值 5.3億美元高額加密貨幣遭竊 虛擬貨幣不再安全?

當地時間2018年1月27日,日本虛擬貨幣交易所的運營巨頭Coincheck在東京召開新聞發布會稱,價值約580億日元(約合34億元人民幣.

1900/1/1 0:00:00
POS:「區塊鏈入門」什么是PoS挖礦?與比特幣的PoW挖礦有什么不同?

前段時間,幣價上漲的時候,寫「區塊鏈入門」這樣的學習知識,看的人很少,所以區塊鏈傳奇就沒有繼續寫這方面的內容。然而,最近一陣的下跌告訴我們一個事實:胡亂買幣就可以賺錢的時代結束了.

1900/1/1 0:00:00
ads