146IEEEJOURNALONSELECTEDAREASINCOMMUNICATIONS,VOL.32,NO.1,JANUARY2014
whentheavailablerunningtimeisverylimitedwhiletheseconddirectionrequiresalongerrunningtime.
(i)We rstcountforeachoftheelementsinthe rstcolumn,thenumberofdistinctentriesinwhichitappears.Asdiscussedearlier,duplicatedentrieshavenoin uenceontheobtainedencodingwidth.Basedonthesenumbersofdistinctappearances,wecancalculateacorrespondingdistribution.Wecanthenencodethis rstcolumnbyaHuffmancoding[24]basedonthisdistribution.Giventhisencodingofthe rstcolumn,wecan ndanencodingschemeforthewholetablebyencodingthesecondcolumnbytheoptimalconditionalencodinggiventheencodingofthe rstcolumn.Anotheroptionforanencodingschemeisobtainedsimilarlybyconsideringthetwocolumnsinareverseorder.ByLemma5andTheorem6,itisguaranteedthatbothoptionsperformatleastaswellasanencodingschemecomposedofthetwoHuffmancodingsforeachofthetwocolumns.Finally,wecanencodethegiventwo-columnFIBtablesbythebestofthetwooptions,i.e.theonethatyieldsasmallerencodingwidth.
(ii)We ndtheencodingschemeobtainedbythealgorithmforgeneralFIBtablespresentedinSectionIII-B.Then,weconsider(independently)asthegiven(constrained)encoding,eachofthetwoencodingsforthetwocolumnscomposingthesolution.Foreachofthem,we ndtheoptimalconditionalencodingoftheothercolumn.Again,byLemma5andTheorem6,itisguaranteedthatbothoptionsperformatleastaswellastheoriginalsolutiongivenbythealgorithm.V.LOWERBOUNDSONTHEOPTIMALENCODINGWIDTH
OFTWO-COLUMNFIBTABLESWewouldlikenowtosuggestadditionallowerboundsontheoptimalencodingwidthofagiventwo-columnFIBtableAasde nedabove.WecancalculatethisboundwithoutsolvingtherelaxedFIBencodingproblem.
LetagainGA=<L+R,E>=<S1+S2,E>with|S|S=n=2W.We rstrecallanadditionalde nitionfrom1|=graph2|theory.
De nition5(EdgeCover).ForanundirectedgraphG=<V, E>,asetofedgesS EiscalledanedgecoverofGif(x,yat)least∈S{x,y}=V,i.e.eachoftheverticesinVisincidentononeoftheedgesinS.
Clearly,GAdoesnotincludeanyisolatedvertices,i.e.anyvertexisconnectedtoatleastonevertexandthevalueitrepresentsappearsinatleastoneoftherowsofA.Therefore,anedgecoveralwaysexists.ForagraphG,wedenotebyρ(G)theedgecoveringnumberofG,i.e.theminimalnumberofedgesinanedgecoverofG.We rstshowasimplepropertyofρ(GA).
Lemma7.Theedgecoveringnumberρ(G)≤2·(n 1).
A)ofG≤ρ(GAsatis esnAProof.Considertheedgesinaminimaledgecoverinanarbi-traryorder.The rstedgecoverstwovertices.Anyadditionaledgeaddsoneortwoverticestothesetofverticescoveredbypreviousedgesintheedgecover.Sincethereare2nverticesinGAthenclearlyn≤ρ(GA)≤2n 1.Toshowthat
ρ(GA)≤2n 2,weshowthattheremustbetwoindependentedgesinGA,i.e.edgesthatdonotshareanyvertex.Consideranedge(s1,i,s2,j)fori,j∈[1,n].IfalltheotheredgesinEshareatleastonevertexwith{(s1,i,s2,k)|k∈[1,n]} thisedge,wemusthavethatE={(s)1,,k,s(s2,j)|k∈[1,n]}.Then,forinstancetheedges(s1,i,s2,k1,k,s2,j)fork=i,jareindependent.Finally,thetwoindependentedgescancon-tributetwocoveredverticestotheedgecoverandnecessarilyρ(GA)≤2n 2=2·(n 1).
Letα=ρ(GA)/n.ThenexttheoremsuggestsalowerboundontheoptimalencodingwidthofAbasedonthevalueofρ(GA).
Theorem8.TheoptimalencodingwidthofAsatis es
OPT(A)≥2·(W+α 1)
α.(20)
Proof.ForGA=<S1+S2,E>,letS EbeanedgecoverofGAofsizeρ(GA)=α·n.LetCtableencodingschemeofAA=(σ.Wewill1,σshow2)beanarbitraryFIBthat
(x,ymax)∈S (σ1(x))+ (σ2(y))≥2·(W+α 1)
α.(21)Thus,sinceS Ethennecessarily
(C2·(W
(x,y)∈E
(σ1(x))+ (σ2(y))≥+α 1)
A)=max α.
(22)Todoso,weconsiderthesubsetoftherowsofArepresentedbytheedge((d11
ρ(GcoverS.LetD=
A),dρ|D|1,d=2),...,(dρ(G1
2(GA)))representtheserows,s.t.A)=α·S1= ρi=1(Gn.ByDe nition5oftheedgecover,
A){di appear1}andSatleast2=ρoncei=1(GA){di
inthe2},i.e.allthe
elementsofS1 rstcolumnoftherowsofDaswellasalltheelementsofSLetX2inthesecond.D(CA)denotethetotalsumofcodewordlengthsoftheelementsinDusingtheencodingCAX D(CA)= =(σ(G1,σ2),s.t.
ρi=1A) (σ1(di1))+ (σ2(di
2)).
BymanipulatingKraft’sinequality,wecanhavethatthetotalsumofcodewordlengthsofthendistinctelementsineachcolumnisatleastn·W.Ofcourse,theotherα·n n=(α 1)·nelementsineachcolumnareencodedinatleastonebiteach.Thus,ρ(X GA)
D(CA)=
(σ1(di1))
+
(σ2(di
2))i=1
ρ(GA)
ρ(GA)
=
(σ
1(di1))
+
(σ2(di2))
i=1
i=1
≥n·W+(α 1)·n+n·W+(α 1)·n=2n·W+2(α 1)·n.
(23)
Werecallthatthetotalnumberofrowsisα·nandisatleastonerowinDthatisencodedinatleast thusX(CthereDα·n
A)
Compressing Video for TaskStream Currently, TaskStream allows users to upload videos up to 100 megabytes in size. Therefore, in order to upload a video...
PPMexe: PPM for Compressing Software Milenko Drini?? and Darko Kirovski? c ? Computer Science Department, UCLA, Los Angeles, USA ? Microsoft Research, ...
Face Fixer Compressing Polygon MeshesFace Fixer Compressing Polygon Meshes隐藏>> UNC Technical Report TR-00-04 Face Fixer: Compressing Polygon Meshes with ...
Compressing Hexahedral Volume Meshes Martin Isenburg? University of North Carolina at Chapel Hill Pierre Alliez? INRIA Sophia-Antipolis Abstract Unstructured ...
Compressing Mixed-State Sources by Sending Classical Information_专业资料。We consider visible compression for discrete memoryless sources of mixed quantum ...
Managing Gigabytes Compressing and Indexing Documents and Images_专业资料。As today’s information explosion generates greater and greater volumes of raw data...
Compressing File Collections with a TSP-Based Approach Dimitre Trendafilov Nasir Memon Torsten Suel Department of Computer and Information Science Technical ...
Compressing cube-connected cycles and butterfly networks We consider the simulation of large cube-connected cycles (CCC) and large butterfly networks (BFN)...
The objectives for compressing DEMs are obtain...更多>>Abstract. Wavelet decomposition is a well-known technique to compress image data. Here, we have ...
Re-Store: A System for Compressing, Browsing, and Searching Large Documents (Extended Abstract of Invited Presentation) Alistair Moffat Raymond Wan ...
热门文档