Representation of linear operators by Gabor multipliers

Size: px
Start display at page:

Download "Representation of linear operators by Gabor multipliers"

Transcription

1 Rprsttio of lir oprtors Rprsttio of lir oprtors by Gbor multiplirs Ptr C Gibso, Michl P Lmourux, Gry F Mrgrv ABSTRACT W cosidr cotiuous vrsio of Gbor multiplirs: oprtors cosistig of short-tim Fourir trsform, followd by multiplictio by distributio o phs spc (clld th Gbor symbol), followd by ivrs short-tim Fourir trsform, llowig diffrt loclizig widows for th forwrd d ivrs trsforms For giv pir of forwrd d ivrs widows, which lir oprtors c b rprstd s Gbor multiplir, d wht is th rltioship btw th (o-clssicl) Koh-Nirbrg symbol of such oprtor d th corrspodig Gbor symbol? Ths qustios r swrd compltly for spcil clss of comptibl widow pirs I dditio, cocrt xmpls r giv of widows tht, with rspct to th rprsttio of lir oprtors, r mor grl th stdrd Gussi widows Th rsults i th ppr hlp to justify tchiqus dvlopd for sismic imgig tht us Gbor multiplirs to rprst osttiory filtrs d wvfild xtrpoltors INTRODUCTION This ppr is cocrd with th rprsttio of lir oprtors by Gbor multiplirs For us Gbor multiplir is oprtor of th form: short-tim Fourir trsform, followd by multiplictio by distributio o phs spc clld th Gbor symbol followd by ivrs short-tim Fourir trsform W xplicitly llow th lysis widow, upo which th forwrd short-tim Fourir trsform is bsd, to b diffrt from th sythsis widow, which srvs s th bsis for th ivrs short-tim Fourir trsform Mor prcisly, w study th rprsttio of grl (i, o-clssicl) Koh- Nirbrg psudodiffrtil oprtors s for xmpl [4] by ms of Gbor multiplirs As is wll-kow, th clss of grl Koh-Nirbrg oprtors is simply ll lir oprtors (sstilly this is vrsio of th Schwrtz krl thorm s (5) i Sctio ) Howvr, our im is to lyz th corrspodc btw th Koh- Nirbrg symbol, which i th grl sttig is rbitrry distributio, d th Gbor symbol, of giv oprtor, providd th lttr xists Th Gbor symbol of oprtor dpds o choic of lysis d sythsis widows; idd for fixd widow pir, thr my b limitd clss of lir oprtors tht my b rprstd s Gbor multiplir As fr s w kow th problm of rprstig Koh-Nirbrg oprtors xctly by (our vrsio of) Gbor multiplirs is w, d i this ppr w cofi ourslvs to th followig two fudmtl qustios CREWES Rsrch Rport Volum 5 (003)

2 Gibso, Lmourux, d Mrgrv For giv pir of lysis d sythsis widows, which Koh-Nirbrg oprtors my b rprstd s Gbor multiplir? Ar thr xplicit formuls rltig th Gbor d Koh-Nirbrg symbols? Much of th ppr rsts o ky tchicl otio of comptibl widow pir For such widows w hv complt swr to both of th bov qustios A commo widow choic for th short-tim Fourir trsform is Gussi, both s lysis d sythsis widows, d this srvs s th prototyp of comptibl widow pir O cocrt rsult tht mrgs from our lysis of comptibl widows is w typ of widow, which w cll xtrm vlu widow lytic, rpidly dcrsig widow, with lytic, rpidly dcrsig Fourir trsform It is mor grl th Gussi i th ss tht th clss of oprtors tht c b rprstd s Gbor multiplir bsd o Gussi widows is propr subst of th clss tht c b rprstd usig xtrm vlu widows Th bsic qustios tht w ddrss hr r motivtd by prticulr pplictio, sismic imgi whr psudodiffrtil oprtors ply importt rol [6,7] d Gbor multiplirs hv b succssfully implmtd s fficit d flxibl mthod of sismic dcovolutio [8] Ths implmttios rst o th fct tht thr xist fficit umricl schms, dscribd i [5], by which to vlut discrtizd vrsios of Gbor multiplirs But somthig tht hs b lckig so fr is thorticl lysis of th rltioship btw th origil oprtors i xtrpoltors or filtrs which r sily xprssd s Koh-Nirbrg oprtors, d th corrspodig cotiuous Gbor multiplirs tht hv b usd to std i for thm Th rsults prstd hr srv to fill i th gp Thr is importt issu cocrig th vlutio of lir oprtors i prcticl pplictios tht w do ot ddrss i th prst ppr, mly, to qutify th dgr of pproximtio to cotiuous Gbor multiplir tht is ttid by discrtizd vrsios of it For rsults i this dirctio, s, for xmpl, Fichtigr d Nowk (003) d [9] Th ppr is orgizd s follows Th mthmticl frmwork i which w formult th problms dscribd bov is md prcis i Sctio, whr w fix th bsic ottiol covtios Thorm of Sctio rfsc-schwrtzkrl givs strightforwrd wy to xprss th Schwrtz krl of lir oprtor i trms of its Gbor symbol, which i tur yilds xprssio for th Koh-Nirbrg symbol i trms of th Gbor symbol Th covrs problm of xprssig th Gbor symbol i trms of th Koh-Nirbrg symbol sms to b much mor difficult, d tks up th bulk of th ppr Sctio 3 provids th bsic motivtio for our ky tchicl dfiitio of comptibl widows i Sctio 3 A ltrt dfiitio of comptibility, provd i Thorm of Sctio 3 to b quivlt to th origil, lds to squc of tchicl rsults i Sctio 3 tht ly th groudwork for our mi Thorms Ths r Thorm 3 d Thorm 4 of Sctio 33 I Sctio 34 w pply Thorm 4 to show tht xtrm vlu widows srv to rprst widr clss of lir oprtors s Gbor multiplirs th is possibl with Gussi widows W discuss brifly i Sctio 35 turl xtsio of th otio of comptibility to sigulr widows Lstly, Sctio 4 provids short summry of th ppr CREWES Rsrch Rport Volum 5 (003)

3 Rprsttio of lir oprtors PRELIMINARIES Nottio d covtios Our ottio is mostly stdrd, with th xcptios tht: (i) w us th vrsio of th Fourir trsform tht hs fctor of π i th xpot, d (ii) w tk tmprd distributios to b cotiuous, cojugt lir, rthr th lir, fuctiols o th spc of Schwrtz clss fuctios Grlly spki w dl with fuctios d distributios o R, whr th vlu of is fixd withi giv cotxt, d R is lwys th domi of itgrtio, which w omit W idict poit i R by pir (x,y) of poits xy, R, d itgrtio ovr R is idictd by pir of itgrl sigs For covit rfrc, w hv compild i Tbl list of som of th fuctio spcs d oprtors tht w us rptdly W work withi th bsic frmwork of L, th spc of cotiuous, lir oprtors L ' : S S, mkig frqut us of th corrspodc btw L d S ' [0] Mor prcisly, ' giv lir oprtor L L, thr is uiqu distributio K = K( L) S, its Schwrtz krl, tht stisfis th qutio K, θ ϕ = Lϕ, θ ϕ, θ S () ' Ad covrsly, giv K S, th qutio () vidtly dtrmis uiqu L L (Hr dots th tsor product: f gxy (, ) = f( xgy ) ( )) Th djoit of lir * oprtor L L is th lir oprtor L L dfid by th qutio * Lϕ, θ = Lθ, ϕ ϕ, θ S () I ordr for th oprtor T ψ, listd i Tbl, to b wll-bhvd, som rstrictios hv to b plcd o ψ ; i this rgrd w itroduc th otio of tmprd chg of vribls, s follows Dfiitio W sy tht smooth, ivrtibl mp m m ψ : R R, * * is tmprd chg of vribls if ch of th oprtors T, T, T, T mps S m ito S m ψ ψ ψ ψ Thr is o tmprd chg of vribls o R whos corrspodig oprtor w ssig spcil ottio: CREWES Rsrch Rport Volum 5 (003) 3

4 Gibso, Lmourux, d Mrgrv T_ = T, whr ψ _( x, y) = ( x, y x) (3) ψ _ Giv rbitrry tmprd distributio oprtor dfid by th formul σ( x, ξ) S, w writ σ ( X, D) for th ' ix ( X D) ( X D) ( x) π ξ σ, : S S ; σ, ϕ = σ( x, ξ) F ϕ( ξ) dξ (4) (Hr X : R R dots th idtity mp, so tht, for xmpl, X ( x) = x D stds for th diffrtil oprtor D = πi ) W rfr to σ ( X, D ) s Koh- Nirbrg psudodiffrtil oprtor; th distributio σ ( x, ξ ) is its Koh-Nirbrg symbol I th cotxt of clssicl psudodiffrtil oprtors, whr th symbol σ ( x, ξ ) is rquird to b smooth d of boudd growth, th itgrl o th right-hd sid of (4) is ihrtly wll-dfid A simpl wy to giv th itgrl umbiguous itrprttio i th prst much mor grl sttig is to dfi σ ( X, D) i trms of its Schwrtz krl: K( σ ( X, D)) = T_F σ (5) ' Sic ch of th oprtors T_ d F crris S bijctivly oto itslf, it is vidt from th rprsttio (5) tht th clss of Koh-Nirbrg psudodiffrtil oprtors o R is idticl with L itslf Amog th vrious wys to rprst lir oprtor, howvr, th Koh-Nirbrg symbol d ccompyig forml rprsttio (4) r of prticulr itrst sic, from th physicl poit of viw, thy r turl both for prtil diffrtil oprtors d for osttiory filtrs [4] I othr words, i pplictios o is somtims giv th Koh-Nirbrg symbol of lir oprtor dirctly Not tht th Koh-Nirbrg symbol σ L d th Schwrtz krl K(L) of lir ' ' oprtor L : S S both blog to S But thr is ss i which ths r distict vrsios of ' S, i tht σ L is distributio o phs spc, R R, whil K(L) is distributio o th cross product R R of th udrlyig spc with itslf Idd, sic it is somtims usful to mk this distictio btw spc d frqucy, w rsrv ξ d η for frqucy vribls, usig othr lttrs (such s x,y,τ,t,v,w) for spcil vribls 4 CREWES Rsrch Rport Volum 5 (003)

5 Rprsttio of lir oprtors Tbl Nottio Fuctio d distributio spcs S Schwrtz clss fuctios ϕ : R ' R S (cojugt lir) tmprd distributios u : S C P C fuctios ϕ : R R such tht ϕ is boudd by polyomil P for vry multi-idx Oprtors Dscriptio Symbol Actio o fuctios Adjoit Compositio with chg of vribls m m ψ : R R Fourir trsform Prtil Fourir trsform Prtil Fourir trsform Modultio Trsltio Multiplictio by ' S m T ϕ ϕ ψ * ψ T F F F S S : ' ' ( S S) S S : ' ' ( S S) S S : M T ' ' ( S S) S S ξ : x ' ' ( S S) S S : ' ' ( S S) N : m ' m ϕ( x) ϕξ ( ) πξ i x = ϕ( x) dx ϕ( x, y) F ϕξ (, y) = dt J T ψ ψ ψ F = F F = F πξ i x = ϕ( x, y) dx ϕ( x, y) Fϕ( x, η) F = F πη i y = ϕ( x, y) dy ϕ M = M πξ ( ) i x x ϕ( x) ξ ξ ϕ() t ϕ( t x) P S ϕ ϕ T = T x x CREWES Rsrch Rport Volum 5 (003) 5

6 Gibso, Lmourux, d Mrgrv Th Schwrtz krl of grl Gbor multiplir If g : R C is msurbl d boudd by polyomil, th th formul V ϕ( x, ξ) = M T ϕ g dfis short-tim Fourir trsform V g : S P with lysis widow g Th bsic thory of short-tim Fourir trsforms is dscribd i [4] For S, th rg of V lis i S ; th djoit V of V is th mp ξ x V : S S ; V u, ϕ = u, V ϕ If th distributio u S hpps to b L fuctio, th V u is giv by th formul Vut () = ux (, ξ ) MT() tdxdξ ξ x Dfiitio Lt g : R C b msurbl d boudd by polyomil, d lt S If i dditio 0, th w sy tht ch of ( ) d (, g) is widow pir o R It is bsic fct bout th short-tim Fourir trsform tht for y widow pir ( ), th mp VV g : S S (6) is th idtity Rcll from Tbl tht w us th symbol by N to dot multiplictio Dfiitio 3 Giv widow pir ( ) o M V N V g R d distributio 6 CREWES Rsrch Rport Volum 5 (003) = Gbor multiplir; w rfr to th distributio s its Gbor symbol S, w cll (Not tht Fichtigr d Nowk [] us th trm short-tim Fourir trsform multiplir for Gbor multiplir bsd o idticl widows ( gg, ), whil i Fichtigr d Nowk (003) Gbor multiplir rfrs to mor grl objct th w hv dfid) A Gbor multiplir, i th ss of Dfiitio 3, is lir oprtor blogig to L Its djoit is lso Gbor multiplir, d th prcis coctio btw th two works out s follows

7 Rprsttio of lir oprtors Propositio 4 For y widow pir ( ) d y distributio S, th djoit of th Gbor multiplir M is, g = M M Roughly spki th dfiig structur of Gbor multiplir ms tht it crris implicit digoliztio o phs spc From th thorticl poit of viw, this fct mks it dsirbl to xprss, if possibl, giv lir oprtor L L s Gbor multiplir, th structur of th oprtor th big codd i its Gbor symbol It is lso dsirbl to xprss oprtor s Gbor multiplir from th poit of viw of pplictios, sic thr xist fst computtiol mthods to vlut discrtizd Gbor multiplirs [5] Th mi problm tht w r cocrd with i th prst ppr is to xprss giv Koh-Nirbrg psudodiffrtil oprtor s Gbor multiplir Bfor cosidrig th issu i dtil, w dl brifly with th covrs problm, of xprssig giv Gbor multiplir s psudodiffrtil oprtor I light of th xprssio (5) for th Schwrtz krl of psudodiffrtil oprtor, th lttr problm is quivlt to computig th Schwrtz krl of Gbor multiplir This turs out to b rltivly strightforwrd, d c b crrid out i full grlity I sttig th bsic rsult w mk us of th followig ottio Lt S : S4 S dot th mp dfid by Th corrspodig djoit is th mp Sρ( x, ξ) = ρ( x, x, ξ, ξ) S : S S ; S u, ϕ = u, Sϕ 4 Thorm A rbitrry Gbor multiplir M hs Schwrtz krl K( M ) = V gs (7) Proof Not tht M ϕ, θ = V, gϕv θ d π ( ) ( ) it ξ πiτ ξ Vg V x, = t g( tx) dt ( ) ( x) d ϕ θ ξ ϕ θ τ τ τ πit ξ πiτ ξ = ϕ () t gt ( x) θτ ( ) ( τxd ) τdt CREWES Rsrch Rport Volum 5 (003) 7

8 Gibso, Lmourux, d Mrgrv Thus, πit ( ) τ ξ ( τ x) g( t x) θ ϕ( τ t) dτdt (8) =, =,, πi( τ, t) ( ξ,ξ) T( xx, ) g( τ t) θ ϕ( τ t) dτdt = V gθ ϕ( x, x, ξ, ξ) M ϕ, θ = SV, gθ ϕ = V gs, θ ϕ Sic grlizd Koh-Nirbrg oprtors compss ll of L, giv rbitrry Gbor multiplir M o R, thr xists distributio σ S such tht σ ( X, D) =M By Thorm, this is quivlt, i trms of Schwrtz krls, to th qutio which c b solvd for σ i trms of to yild T_F σ = V gs, (9) σ = F T V g S (0) I cotrst, thr is o obvious wy to solv qutio (9) for i trms of σ, sic ithr of th oprtors V or S g is ijctiv Idd, dtrmiig th Gbor symbol tht corrspods to giv Koh-Nirbrg symbol sms to b much mor itrict problm Idticl Gussi widows Lt us spciliz to th cs of widow pir cosistig of idticl Gussis, gt () = () t = π t, whr w us th ottio t = t t for t R Th spcil structur of ths widows ms tht w c obti ltrt formul for th Schwrtz krl of y Gbor multiplir bsd o thm Th ky poit i our proof of Thorm, t which w c tk dvtg of our prst choic of Gussi widows, is th itgrl (8) Th itgrd ivolvs th product 8 CREWES Rsrch Rport Volum 5 (003)

9 Rprsttio of lir oprtors π( τx) π( tx) ( τ xgt ) ( x) = Thus, i trms of th tmprd chg of vribls w s tht ( τ x) gt ( x) hs th form π( τ+ t x) π ( tτ) = ( vw, ) = (, t τ ), τ + t ( τ x) gt ( x) = G( v xg ) ( w), () whr V ϕ g V () v π G v θ, w obti = d ( ) w G w = π Substitutig () ito th xprssio (8) for πit ( τ) ξ VgϕV θ = G( vx) G( w) θ ϕ( τ, t) dτdt = G ( vx) G ( w) θ ϕ ψ ( v, w) dtj dvdw πiw ξ ψ T =, ψθ ϕ, πiw ξ ( ) T ( ) ( x, 0) G G vw vwdvdw T πiw ξ ( ) T ( ) = ( x, 0) G G v, w ψθ ϕ v, w dv dw ( G = F G T ψθ ϕ ), () whr G () v G ( v) = d dots covolutio with rspct to th first vribls Rplcig F with F, d crryig F through th covolutio, w obti from () tht VgϕVθ = F ( G G FT ψθ ϕ) Procdig s i th proof of Thorm, this lds to th followig formul for th Schwrtz krl of M : K ( M ) = T N ψf F (3) G G Rcllig tht gt () = () t = π t, () v π G v =, d ( ) w G w = π, w hv Sic πt / = dt = ( ) πη π w G G η, w =, th formul (3) simplifis s i th xt propositio CREWES Rsrch Rport Volum 5 (003) 9

10 Gibso, Lmourux, d Mrgrv Propositio 5 Lt S gt () = () t = π t b idticl Gussis o R Th for y, th Schwrtz krl of M is K( M ) = T F N F, (4) g, ψ G whr G : R R is th Gussi τ + t vribls ψ ( τ, t) = (, t τ) (, w ) G( w) π η η, = d ψ is th tmprd chg of Ech of th oprtors T, F, N d F occurrig i th xprssio (4) for th ψ G Schwrtz krl of M is ijctiv Thrfor if M = σ ( X, D), w c, by comprig th Schwrtz krls of th two oprtors, xprss th Gbor symbol of M i trms of σ s (, y ) i y N whr H( y) π η π η = F F σ, η, = (5) H Morovr, for giv σ S, th oprtor σ ( X, D) c b xprssd s Gbor multiplir bsd o idticl Gussi widows oly if th xprssio (5) yilds wlldfid tmprd distributio (Not tht (5) cot b writt s covolutio, bcus th rpidly icrsig fuctio H dos ot hv Fourir trsform) Ar thr widow pirs ( ) othr th idticl Gussis for which o c obti rsults logous to ths? Th purpos of this ppr is to dscrib clss of widow pirs for which this is possibl MAIN RESULTS Dfiitio d xmpls of comptibl widow pirs Th xprssio (4) for th Schwrtz krl of Gbor multiplir bsd o idticl Gussis is th bsis for th formul (5) With y to obtiig mor grl rsults log ths lis, w ivstigt th clss of ll widow pirs for which formul logous to (4) xists I this rgrd th sstil proprty of Gussi is th fctoriztio rul (), so w focus o th clss of ll widow pirs tht oby such rul This lds dirctly to th followig ky tchicl otio Dfiitio 6 W sy tht widow pir ( ) o R is comptibl, d tht g d r comptibl widows, if thr xist fuctios G, G : R C d tmprd chg of vribls ( vw, ) = ψ ( τ, t) with w = t tu such tht for vry τ,, tx R, ( τ x) gt ( x) = G( v xg ) ( w) 0 CREWES Rsrch Rport Volum 5 (003)

11 Rprsttio of lir oprtors Not tht our drivtio of formul (3) i Sctio is vlid for rbitrry comptibl widows s w hv just dfid thm Hc w immditly hv th followig rsult Propositio 7 Lt ( ) b comptibl widow pir o Schwrtz krl of M is K ( M ) = T N ψf F, G G whr G, G d ψ r rltd to ( ) s i Dfiitio 6 R Th for y S If ( ) is comptibl th so is (, g) Th prcis coctio btw th ccompyig sts of fuctios s sttd blow my b sily vrifid Hr w writ ψ : R R for th chg of vribls ψ ( x, y) = ( y, x), d w us th ottio F ( x) = F( x) Propositio 8 Lt ( ) b comptibl widow pir o ( τ x) gt ( x) = G( v xg ) ( w), R, so tht i ccordc with Dfiitio 6 Th th rvrs widow pir (, g) is lso comptibl d stisfis th qutio, th whr g( τ x) ( t x) = G ( v x) G ( w), G = G, G = G, d ( v, w) = ( v ψ w), Tbl lists svrl fmilis of comptibl widow pirs, logsid thir corrspodig fctoriztios d chgs of vribls, s prscribd by Dfiitio 6 Th clss of comptibl widow pirs is sily s to b ivrit with rspct to umbr of lmtry oprtios: rscli trsltio d modultio of ithr widow, d lir chg of vribls pplid simultously to both widows This obsrvtio shows tht th list i Tbl is ot complt O th othr hd, th widow pirs listd i Tbl r rsobly rprsttiv For xmpl, it turs out tht vry ogtiv comptibl widow pir o R c b obtid by trslti dilti rvrsig or rsclig th giv xmpls, or by rvrsig th ordr of widow pir obtid this wy This is rthr ivolvd tchicl rsult tht will b prstd i forthcomig ppr [3] CREWES Rsrch Rport Volum 5 (003)

12 Gibso, Lmourux, d Mrgrv Tbl Exmpls of comptibl widows Th prmtrs m µ k κ 5) r positiv sclrs Th widow pirs of lis (-3) r dfid o widow pirs of lis (4,5) r dfid o R oly Comptibl widow pirs ( ) # gt () ( τ ) ( vw, ) = ψ ( τ, t) G () G ( ) S ( τ, t τ ) () v πξ i t S ( τ, t τ ) i v () v πξ i w 3 mt µτ mt+ µτ m ( m+µ ) v w ( m+ µ, t τ m ) µ + µ t 4 t τ τ log ( τt t ) + v +, tτ v cosh( w/ ) t 5 kt τ κτ ( ( ) ) ( ) ( τt + t+ log ), tτ ( ) ( k+ κ ) v v,,,, pprig i lis (3- ( ) + R for y ; th / k κ κ w k + w k+ κ k+ κ Th xmpl of li (), Tbl, which stisfis th dfiitio of comptibility i trivil wy, is ot dirctly of prcticl itrst Of cours th Gussi widows of li (3) r th prototyp for th otio of comptibility, d thy r stdrd widow choic for th short-tim Fourir trsform Th most itrstig xmpls i Tbl r th widows of li (4) d thir grliztios i li (5) Ths wr discovrd usig th ltrt chrctriztio of comptibility dvlopd i th xt sctio of this ppr; thir full drivtio will b prstd i [3] Lik Gussis, ths widows hv fudmtl rol i probbility thory d ris i coctio to vrit of th ctrl limit thorm Mor prcisly, lt E dot th fuctio g of li (4) i th spcil cs = : t E() t = Proprly scld ffi trsformtios of E, of th form ( t E b ) b, r th dsity fuctios of so-clld xtrm vlu probbility distributios, dscribd by Fishr d Tipptt i [] Bsd o this, w rfr to th widows of lis (4, 5) s xtrm vlu widows s Figur t CREWES Rsrch Rport Volum 5 (003)

13 Rprsttio of lir oprtors t t FIG Extrm vlu widows Th fuctio f () t =, bov, d its rvrsl f ( t), blow, r plottd for = 0 Not th f ( t) is clos to big cusl, i zro for gtiv t vlus of t This is bcus th doubl xpotil tds to th Hvisid fuctio with icrsig Not tht th Fourir trsform of xtrm vlu widow is proportiol to th gmm fuctio, rstrictd to li of costt rl prt: t kt k π i F ( ξ ) = Γ ξ Thus th Fourir trsform of xtrm vlu distributio is lytic, vr zro, d rpidly dcrsig I Sctio 34, w show tht xtrm vlu widows hv crti dvtg ovr Gussis with rspct to th rprsttio of lir oprtors A ltrt chrctriztio of comptibility Hr w dvlop ltrt chrctriztio of comptibility tht mks it sir to driv formuls d to prov som tchicl fcts tht will b dd i subsqut CREWES Rsrch Rport Volum 5 (003) 3

14 Gibso, Lmourux, d Mrgrv sctios Th rgumts giv hr r purly tchicl; thy build towrd our mi rsults which r prstd i th followig Sctio 33 Lt ( ) b comptibl pir o R with corrspodig fuctios G, G d tmprd chg of vribls ( vw, ) = ψ ( τ, t), s stipultd i Dfiitio 6 It c b dducd from th bsic qutio, ( τ xgt ) ( x) = G( vxg ) ( w) τ, tx, R, tht thr r svrl simplifyig ssumptios to b md cocrig G, G d ψ without y icumbt loss of grlity (6) Firstly, ot tht G (( v τ, t) x) G ( t τ) = ( τ x) g( t x) = (( τ t) ( xt)) g(0 ( x t)) = G (( v τ t, 0) ( xt)) G ( t τ ) = G (( v τ t, 0) + t x) G ( w) (7) This lds to th first of our simplifyig propositios Propositio 9 It my b ssumd without loss of grlity tht v( τ, t) = v( τ, t 0) +, t τ t R (8) Proof Suppos tht (8) fils to hold, d cosidr th w fuctio v, dfid by v ( τ, t) = v( τ, t 0) +, t τ t R Bsd o th fct tht ψ = ( vw, ) is tmprd chg of vribls, it my b vrifid tht th chg of vribls ψ = ( v, w) is tmprd lso Ad by costructio, Now, pplyig th idtity (7) yilds v ( τ, t) = v ( τ, t 0) +, t τ t R G ( v x) G ( w) = G ( v( τ t, 0) + tx) G ( w) = G( v x) G( w) Thrfor th dfiig qutio (6) rmis stisfid if w rplc ψ by th w chg of vribls ψ 4 CREWES Rsrch Rport Volum 5 (003)

15 Rprsttio of lir oprtors Propositio 0 It my b ssumd without loss of grlity tht v (0, 0) = 0 Proof Th qutio tht dfis comptibility, (6), rmis stisfid if G is rplcd by d v is rplcd by G ( y) = G ( y+ v(0, 0)) v ( τ, t) = v( τ, t) v(0, 0) Th modifid chg of vribls ψ = ( v, w) is tmprd, d by costructio v (0, 0) = 0 Propositio Without loss of grlity, G (0) = d G Proof Substitutig w = 0 (i, t = τ ) d x = 0 = g ito th qutio (6) yilds () tgt () = G( vtt (, )) G(0) (9) Sic 0 = g, it follows from (9) tht tht G (0) 0 Th qutio (6) vidtly rmis stisfid if G is rplcd by G (0) G d G is rplcd by G(0) G, so w my ssum G (0) = Assumig i dditio tht (8) holds d tht v (0, 0) = 0, w hv tht vtt (, ) = v(0, 0) + t= t It th follows from qutio (9) tht G () t = () t g () t It will strmli subsqut rgumts to ssum tht G, G d ψ hv th forms giv i Propositios 9, 0, d W ow itroduc th otio of product-trsltio commuttivity for widow pir Dfiitio Lt ( ) b widow pir o R W sy tht ( ) is producttrsltio commuttiv, or tht g d r product-trsltio commuttiv, if thr xist fuctios C : R C d ζ : R R such tht for vry w R, gt = C( w) T ( g ), (0) w ζ ( w) d such tht ( τ, t) ( tζ( t τ), tτ) is tmprd chg of vribls Although it is ot immditly obvious, product-trsltio commuttivity of widow pir is quivlt to comptibility This fct fcilitts th lysis of comptibility CREWES Rsrch Rport Volum 5 (003) 5

16 Gibso, Lmourux, d Mrgrv Thorm A widow pir ( ) is product-trsltio commuttiv if d oly if it is comptibl Morovr, if widow pir ( ) hs this proprty th, without loss of grlity, th trms occurrig i th rspctiv dfiig qutios, gt = C( w) T ( g )d ( τ x) g( t x) = G ( v x) G ( w), w ζ ( w) my b ssumd to b rltd by th formuls G = C, v( τ, t) = tζ( t τ)d ζ( y) =v( y, 0) Proof W prov th quivlc of Dfiitios 6 d ; th sttd formuls mrg i th cours of th proof Suppos tht ( ) is comptibl, with ssocitd fuctios G, G d ψ tht stisfy th proprtis i Propositios 9, 0, d Th bsic qutio of Dfiitio 6, qutio (6), c b writt symboliclly s T g = T T G G ( xx, ) ψ ( x, 0) = T g T( x, x) ψt( x, 0) G G g( τ, t) = G ( v( τ + x, t+ x) x) G ( w) () Dfi ζ ( y) =v( y, 0), so tht substitutig x = t ito () yilds g( τ, t) = G (( v τ t, 0) + t) G ( w) = ζ G ( t ( w)) G ( w) = ζ ζ gt ( ( w)) ( t ( w)) G( w) (by Propositio ) () Th idtity w= t τ implis tht g( τ, t) = gtw ( t) Thrfor () is quivlt to th qutio gt () t = G ( w) T ( g )() t (3) w ζ ( w) Not tht v( τ, t) = v( τ t, 0) + t = tζ( t τ), so th tmprd chg of vribls ψ = ( vw, ) is prcisly th mp ( τ, t) ( tζ( t τ), tτ) Togthr with (3), this vrifis tht th pir ( ) coforms to Dfiitio of product-trsltio commuttivity, with C = G Covrsly, suppos tht th widow pir ( ) is product-trsltio commuttiv, with ssocitd fuctios C, ζ s i Dfiitio St v( τ, t) = tζ( t τ ), so tht 6 CREWES Rsrch Rport Volum 5 (003)

17 Rprsttio of lir oprtors (( v τ, t), w( τ, t)) = ( tζ( t τ), t τ) is tmprd chg of vribls As bfor, g( τ, t) = gt ( t), so by dfiitio, w g( τ, t) = ( tζ( w)) g( tζ( w)) C( w) T( xx, ) g( τ, t) = ( tζ( w) x) g( tζ( w) x) C( w) = ( g)( v x) C( w) Thus ( ) coforms to Dfiitio 6, with ssocitd fuctios G ψ = ( vw, ) = g, G = C d W r ow i positio to driv two dditiol formuls cocrig comptibl widow pirs Rcll tht X : R R dots th idtity mp; X k dots its k -th compot W writ ( gx ) for th vctor ( gx ) ( ),, gx, d similrly ( ) ( ) gx, = gx,,, gx, A tild ovr fuctio idicts rflctio i th rgumt: F ( y) = F( y) Propositio 3 If ( ) is comptibl pir th, without loss of grlity, th ssocitd fuctio G is giv by th formul G g = (4) Ad th chg of vribls ψ c b xprssd s ψ ( τ, t) = ( tζ ( t τ), t τ), whr th rstrictio of ζ to poits w t which G ( ) 0 w is giv by th formul ( gx ) gx, ζ = g (5) Proof By Thorm, g d stisfy th qutio Itgrtig both sids ovr gt = G ( w) T ( g ) w ζ ( w) R d rrrgig trms yilds CREWES Rsrch Rport Volum 5 (003) 7

18 Gibso, Lmourux, d Mrgrv G ( w) = T gt ζ ( w) w ( g ) g ( w) = g g ( w) = Obsrv tht Also, ( gx ) ( w) = XgT = XG ( w) T ( g ) w ζ ( w) = G ( w) ( X + ζ ( w)) g g = G ( w), so, providd G ( ) 0 w, th right-hd sid of (5) bcoms G ( w) ( X + ζ( w)) g gx, gx, = ζ( w) + c gx, = ζ( w) G ( w) It turs out tht G c vr b zro, so tht i fct th formul (5) is vlid vrywhr, but w will ot prov this i th prst ppr (s [3]) Nxt w prov som tchicl rsults, i ticiptio of th comig sctio, whr w rlt th Gbor symbols of Gbor multiplirs bsd o comptibl widows to thir Koh-Nirbrg symbols Propositio 4 Lt ( ) b comptibl widow pir, with ssocitd fuctios G, G Th without loss of grlity G ˆ ˆ G = ( g ) ( g ) Proof This is just pplictio of Propositios d 3 8 CREWES Rsrch Rport Volum 5 (003)

19 Rprsttio of lir oprtors Rcll tht ψ dots th prticulr chg of vribls ψ ( τ, t) = ( τ, t τ) d tht T_ dots th corrspodig oprtor T ψ Lmm 5 Lt ( ) b comptibl widow pir, with ssocitd chg of vribls ψ ( τ, t) = ( tζ ( t τ), t τ) (s i Thorm ) Th ψ ( τ, t) = ( tζ ( t τ), t τ) Proof Writ ( f( xy) f( xy)) ψ ( xy),,, =, Th, =, =, ( x y) ψ ψ ( x y) ( f ζ( f f) f f) Combiig th bov xprssios for x d y yilds tht Thus, f = x y+ ζ ( y) ψ ψ ψ ζ ( x, y) = ( f, f ) = ( f, f f ) = ( x+ ( y) y, y) Propositio 6 Lt ψ ( τ, t) = ( tζ ( t τ), t τ) b th chg of vribls ssocitd to i comptibl widow pir ( ) Th FT T πϕ F, whr = ψ ϕ( η, y) = η ( y ζ( y)) Proof First ot tht T T T ψ ψ Lt F( τ, y) S b dummy fuctio, with τ ψ = srvig s Fourir dul vribl to η Th ( FT ) πτη i ψ ψf ( η, y) = Tψ ψf( τ, y) dτ πτη i F( τ ζ( y) y y) dτ (by Lmm 5) = +, =, πη i ( yζ( y)) πτη i F( τ y) d F ( ) πη i ( yζ( y)) = F η, y τ By dsity of S i S d cotiuity of th oprtors i qustio, it follows tht FT T = i πϕ F ψ CREWES Rsrch Rport Volum 5 (003) 9

20 Gibso, Lmourux, d Mrgrv Rltio btw th Gbor d Koh-Nirbrg symbols I Sctio 3 w drivd xplicit formul for th Gbor symbol of Gussi Gbor multiplir i trms of its Koh-Nirbrg symbol, d th usd this to costri th clss of Koh-Nirbrg oprtors tht c b rlizd s such Gbor multiplir W ow prst th corrspodig rsults i much mor grl stti mly, for rbitrry comptibl widows Thorm 3 Lt S d lt ( ) b comptibl widow pir o R Th M = σ ( X, D) if d oly if th symbols d σ r rltd by th qutio ˆ πϕ i ( g ˆ ) ( g ) F = F σ, (6) whr th phs fuctio ϕ is giv i trms of ( ) by th formul ( gx ) ( y) gx, ϕη (, y) = η y g ( y) Proof Comprig th Schwrtz krl of M, giv i Propositio 7, to th Schwrtz krl of σ ( X, D), giv i (5), w hv TF ψ N F = TF G σ G G G F = FT T ψ F σ (7) Applyig Propositio 4 to th lft-hd sid of (7), d Propositio 6 togthr with th formul (5) of Propositio 3 to th right-hd sid, yilds th dsird qutio It is prhps clrr to formult qutio (6) with th trms tht dpd o th widows combid ito sigl fuctio F, s follows: πϕ i FF= F σ, whr F = ( gˆ ˆ ) ( g ) (8) Of cours o c solv xplicitly for th Gbor symbol to yild = N/ F σ, F F (9) with F s i (8) For th cs whr g d r comptibl, qutio (9) costituts th compio rsult to qutio (0) i Sctio Th xprssios 0 CREWES Rsrch Rport Volum 5 (003)

21 Rprsttio of lir oprtors ( gˆ ˆ ) ( g )d πiϕ corrspodig to ch of th comptibl pirs i Tbl r workd out i Tbl 3, whr th cosqut rltios btw Gbor d Koh-Nirbrg symbols r listd Th followig chrctriztio of oprtors tht c b rprstd s Gbor multiplir bsd o giv widow pir rsts sstilly o Thorm 3 Tbl 3 Compriso of Gbor d Koh-Nirbrg symbols for comptibl widows Not tht, s i Tbl, m, µ,, k, κ dot positiv sclrs, d th widows of lis (4,5) r dfid o R oly Gbor vs Koh-Nirbrg symbols for comptibl widow pirs ( ) # gt () ( τ ) Rltio btw F ( η, y) d F σ ( η, y) S ( η) F = F σ π iξ t S (0) πξ i y ( ) ( ) η + ξ F = F σ ξ 3 mt µτ mµ π η y πm iη y m+ µ m+ µ m+ µ t 4 t τ τ t 5 kt τ κτ Γ F= F σ / y πη i πi y ( η Γ ) y ( ) σ / F = + F Γ (/ ) ( + ) + κ π k+ κ ( ) ( ) k+ p κ y k y k+ κ k+ κ + πη i y k i η Γ F= ( + ) F σ Thorm 4 Lt σ S Th Koh-Nirbrg oprtor σ ( X, D) c b rprstd s Gbor multiplir bsd o giv comptibl widow pir ( ) if d oly if σ GS, or quivltly, σ H S, whr G = ( gˆ ˆ ) ( g ), H = ( g) ( gˆ ˆ ) Proof Sic th chg of vribls ( τ, t) ( tζ( t τ ), t) ssocitd to ( ) is rquird to b tmprd, th fuctio hs th proprty tht h( η, y) = = πϕη i (, y) πη i ( yζ( y)) hs = (/ h) S = S Now, for fixd oly if σ S, th qutio (6) of Thorm 3 holds for som S if d CREWES Rsrch Rport Volum 5 (003)

22 Gibso, Lmourux, d Mrgrv ( gˆ ˆ ) ( g ) ˆ Fσ S ˆ ( ) ( ) (( ) ( ˆ ˆ = g g S σ g g) ) S h W c rd off from this rsult som bsic huristic pricipls cocrig th xistc of Gbor multiplir bsd o giv widows tht rprsts giv Koh- Nirbrg oprtor, s follows I ordr to rprst σ ( X, D) s Gbor multiplir bsd o th comptibl widow pir ( ), it hs to b th cs tht ση, ( y) dcys s fst s gˆ ˆ i th vribls η, d s fst s g i th vribls y Th symbol σ ( x, ξ ) itslf hs to b s smooth s g i th vribls x, d s smooth s ˆˆ g i th vribls ξ Thorm 4 shows tht for Gbor multiplirs bsd o giv widow pir ( ) to rprst wid clss of lir oprtors, th fuctio H should pproximt dlt fuctio, so tht H S S Not tht th scop of idticl widows g = is limitd by th ucrtity pricipl: it is ot possibl for g d ˆˆ g to both b highly loclizd, so th symbol σ ( x, ξ ) of oprtor rprstbl usig idticl widows must b slowly vryig i ithr x or ξ This costrit c b voidd by usig distict widows, providd o of is highly loclizd whil th Fourir trsform of th othr is highly loclizd I kpig with this ssrtio, w will s i Sctio 35 tht mximum grlity is chivd with th sigulr widow pir ( ) = (, δ ) d its rvrsl ( δ, ) Grlity of Gussi vrsus xtrm vlu widows With rspct to th rliztio of lir oprtors s Gbor multiplirs, th grlity of giv widow pir ( ) is rprstd by th st Rg (, ) = σ S S such tht σ( XD, ) = M Ad so, giv two widow pirs ( ), ( ) o grl th ( ) if R, w sy tht ( g, ) is mor or quivltly, if Rg (, ) Rg (, ), Rg (, ) Rg (, ) I trms to this ottio, Thorm 4 sys tht for comptibl widow pir o R, CREWES Rsrch Rport Volum 5 (003)

23 Rprsttio of lir oprtors ( ) ( ) ( ˆ ˆ) d ( ) ( ˆ R = g g S R = g ˆ ) ( g ) S, which llows us to compr th grlity of y two comptibl widow pirs For xmpl, th xt rsult shows tht xtrm vlu widows r mor grl th Gussis Propositio 7 Lt, m, µ > 0 b fixd O ( () ()) is mor grl th th widow pir R, th widow pir t t g t, t =, ( () ()) t t mt t g t, t = µ, Proof W crry out som computtios so s to b bl to pply Thorm 4 Writ From Tbl 3, K = ( g ) ( g ), K = ( g ) ( g ) π i ( η) πη i + y K ( η y) ( ) y g / ( ), Γ / + Not tht KS = KS, whr Ad KS = K S, whr, = Γ, mµ π η y m+ µ m+ µ K ( η, y) = πη i + η Γ (/ ) g, K (, y) = K ( η, y) π i ( η) =Γ y y / ( + ) K ( η, y) = K ( η, y) m π η y m+ µ m+ µ = µ W clim tht K / K P, from which it follows tht ( K / K ) S S To vrify th clim, it suffics to comput th rt of dcy of K d to s tht it is lss th tht of K By Stirlig s pproximtio this works out to b CREWES Rsrch Rport Volum 5 (003) 3

24 Gibso, Lmourux, d Mrgrv ( ) η ( ) y K η, y η π +, which is slowr th th Gussi dcy of K Now, lt ρ S b rbitrry, d cosidr th distributio K ρ : (( ) ρ) K ρ = K K / K K S, which provs tht KS KS, or quivltly KS KS By Thorm 4, this shows tht ( ) is mor grl th ( ) Sigulr widows Th clss of widow pirs stblishd i Dfiitio is usful bcus Gbor multiplirs bsd o thm r wll-dfid irrspctiv of th Gbor symbol W wr thus fr to crry out somwht grl lysis, ucumbrd by qustios of itrprttio or scop of vlidity of formuls But if o is itrstd, ot i grl lysis, but rthr i studyig prticulr oprtors such s, sy, spcific typ of psudodiffrtil oprtor, th thr is good rso to cosidr widows of mor grl typ This coms of cours t pric Gbor multiplirs bsd o mor grl widows my b wll-dfid oly for limitd clss of Gbor symbols, d o hs to crry out som sort of d hoc lysis to dtrmi this clss W will cosidr fw prticulr xmpls of such sigulr widows To bgi, w stblish th rlvt dfiitios Dfiitio 8 If S r such tht o of th xprssios or, g is wlldfid d o-zro, but ( ) is ot widow pir i th ss of Dfiitio, th w cll ( ) sigulr widow pir o R Dfiitio 9 W sy tht sigulr widow pir ( ) o R is comptibl, d tht g d r sigulr comptibl widows, if thr xist distributios G, G S d chg of vribls ( vw, ) = ψ ( τ, t) (ot cssrily tmprd) with w= t τ such tht for vry x R, th qutio holds i th ss of distributios T g = T T G G ( xx, ) ψ ( x, 0) Prhps th simplst wy to obti sigulr comptibl widows is s limits of rgulr comptibl widows For istc, if k δ is pproximt idtity, th th comptibl pirs (, k ), which hv cosistt chg of vribls, td to th limit (, δ ) Th pir (, δ ) is idd sigulr comptibl, with ccompyig fuctios G = δ, G = d chg of vribls ψ ( τ, t) = ( τ, t τ) Not tht, for vry ϕ, θ S, 4 CREWES Rsrch Rport Volum 5 (003)

25 Rprsttio of lir oprtors Thus πix ξ δ V ϕ( x, ξ) = ϕ( ξ), V θ( x, ξ) = θ( x) δ πix ξ M, ϕ, θ =, θ ϕ (30) Sic ix π ξ θ ϕ S, th qutio (30) shows tht th Gbor multiplir M is wlldfid for vry S,δ Th Schwrtz krl of It follows tht M my b computd dirctly from (30):,δ ix θ ϕ = F θ ϕ πix ξ π ξ = F T θ ϕ ( TF ) = θ ϕ K M = T F (3),δ ( ) (W could hv ltrtivly pplid th formul of Propositio 7 to obti th sm rsult) Th xprssio (3) is of cours th Schwrtz krl of th Koh-Nirbrg psudodiffrtil oprtor hvig symbol Tht is, chgig ottio slightly so tht σ dots th Gbor symbol i plc of, w hv: Propositio 0 For vry σ S, M = σ ( X, D), δ σ This rsult is of thorticl importc i tht it shows Gbor multiplirs to b compltly grl, t lst oc w llow sigulr widows Applyig Propositio 8 d th djoit formul of Propositio 4 yilds tht M M δ,, δ = for y S Thus ( δ, ) too is sigulr comptibl pir which givs ris to wlldfid Gbor multiplir irrspctiv of th symbol Although thy r gui xmpls of sigulr comptibl widows, th pirs (, δ ) d ( dlt, ) do ot rlly giv ythig w, sic Gbor multiplirs bsd o thm r just Koh-Nirbrg oprtors (or thir djoits) i disguis O th othr hd, if w lt td to i th comptibl pir o t kt, τ κτ R, w obti somthig w, mly th pir CREWES Rsrch Rport Volum 5 (003) 5

26 Gibso, Lmourux, d Mrgrv kt ( gt ( ), τ ( )) = H( t ) H( τ) κτ,, (3) whr H dots th Hvisid jump fuctio d k, κ > 0 r costts Th ssocitd ( ) fuctios () ( ) k+κ v G v = H v, kw ( kh ( w) κ H ( w)) w w 0 G ( w) = =, κ w w> 0 d th chg of vribls ( vw, ) = ψ ( τ, t) = (mx{ τ, t}, t τ), vrify tht (3) coforms to th dfiitio of sigulr comptibl widow pir By Propositio 7 th Gbor multiplir M bsd o th widows (3) hs krl K ( g, M ) = T N ψf F, (33) G G providd th lttr is wll-dfid A short clcultio yilds tht k + κ ( kh ( y) κ H ( y)) y G G ( η, y) = k+ κ πiη Not tht ( kh ( y) κ H ( y)) y is ot diffrtibl t y = 0, which implis tht if ( η, y) is sigulr t som poit ( η, 0) th th usul distributiol clculus brks dow for th right-hd sid of (33) I this cs th itrprttio of g, M rquirs dditiol clrifictio (which i prctic c b quit strightforwrd) W hv lrdy poitd out tht for y S, th pir (, ) is comptibl, d hc so is its rvrsl (, ) I dimsio =, limitig cs of th lttr is th pir (, ), ( t) = H( t) t (34) Th tructd xpotil i th pir (34) is dvtgous for coupl of rsos Not oly is it mbl to xplicit clcultio, but lso th sigulrity t 0 rdrs th pir (, ) mor grl th if wr Schwrtz clss fuctio, s w will ow show Not first tht (34) is sigulr comptibl with ssocitd fuctios G =, G =, ψτ (, t) = ( t, t τ) Ad th Gbor multiplir, M mks ss for rbitrry symbol S Furthrmor,, G πiη πϕ i ( y) G η, =, FT T ψ F = F, whr ϕ( η, y) = ηy Thus if Koh-Nirbrg oprtor σ ( X, D) c b rprstd s, Gbor multiplir M bsd o th widows (34), th by (7) th Gbor d Koh- Nirbrg symbols r rltd by th qutio 6 CREWES Rsrch Rport Volum 5 (003)

27 Rprsttio of lir oprtors F η, y = F σ η, y πη i πη i y ( ) ( ), ξ = πξ σ (35) ix x ( ) ( i ) π ξ Th qutio (35) imposs o rstrictios o σ for th xistc of corrspodig Gbor symbol I othr words, th widow pir (34) is compltly grl: t RH ( ( t ), ) = S SUMMARY Th strtig poit for th prst ppr ws th two-fold prmis tht (i) i prctic o oft coutrs lir oprtors i Koh-Nirbrg form, σ ( X, D), d tht (ii) it is of itrst both from th umricl d lyticl poits of viw to xprss such oprtor s Gbor multiplir M = V N V, for som choic of loclizig widows g W hv dtrmid clss of widow pirs, mly, comptibl pirs d sigulr comptibl pirs, for which thr xists xplicit formul by which to comput th Gbor symbol i trms of th giv Koh-Nirbrg symbol σ Mor prcisly, th rtio σ / of th Fourir trsforms of th symbols is giv i trms of th widows by th xprssio whr th phs fuctio ϕ is πϕ i ( gˆ ˆ ) ( g ), (36) ( gx ) ( y) gx, ϕη (, y) = η y g ( y) g (37), Extrm vlu widows r prticulrly itrstig xmpl of comptibl widows which r pproximtly cusl (or rvrs cusl), d r mor grl th Gussis with rspct to th rprsttio of lir oprtors Th clss of sigulr comptibl widows icluds widow pirs i trms of which vry lir oprtor my b rprstd s Gbor multiplir Withi this frmwork, th Koh-Nirbrg formultio itslf is xmpl of Gbor multiplir, bsd o th sigulr widow pir (, δ ) Ths rsults offr th prospct of ivstigtig (or vlutig) prticulr lir oprtors, for xmpl prtil diffrtil oprtors or osttiory filtrs, i trms of thir myrid rprsttios s Gbor multiplirs I coclusio w mtio coupl of op problms stmmig from th prst ppr Cocrig th formul (36) for th rtio of th trsformd symbols, it is kow to b vlid for comptibl d sigulr comptibl widows But r thr othr, ocomptibl widows for which it is lso vlid? Mor grlly, is thr brodr clss of widow pirs, for which formuls logous to (36) d (37) c b obtid? CREWES Rsrch Rport Volum 5 (003) 7

28 Gibso, Lmourux, d Mrgrv REFERENCES [] Hs G Fichtigr d Krysztof Nowk, A first survy of Gbor multiplirs, i Advcs i Gbor lysis, Hs G Fichtigr d Thoms Strohmr, ditors, Applid d Numricl Hrmoic Alysis, Birkhäusr, Bosto, 003 [] R A Fishr d L H C Tipptt, Limitig forms of th frqucy distributio of th lrgst or smllst mmbr of smpl, Procdigs of th Cmbridg Philosophicl Socity, XXIV Pt (98), [3] Ptr C Gibso d Ptr Zizlr, Comptibl widows i Gbor lysis (workig titl), i prprtio [4] Krlhiz Gröchi Foudtios of tim-frqucy lysis, Applid d Numricl Hrmoic Alysis, Birkhäusr, Bosto, 00 [5] Michl P Lmourux, Ptr C Gibso, Jff P Grossm d Gry F Mrgrv, A fst, discrt Gbor trsform vi prtitio of uity, 35pp, submittd to Th Jourl of Fourir Alysis d Applictios [6] Gry F Mrgrv d Robrt J Frguso, Wvfild xtrpoltio by osttiory phs shift, Gophysics, 64 (999), [7] Gry F Mrgrv, Robrt J Frguso, d Michl P Lmourux, 00, Approximt Fourir Itgrl Wvfild Extrpoltors for Htrogous, Aisotropic Mdi, Cdi Applid Mthmtics Qurtrly, i prss [8] Gry F Mrgrv, Michl P Lmourux, Jff P Grossm, d V Iliscu, Gbor dcovolutio of sismic dt for sourc wvform d Q corrctio, i 7d A Itrt Mt Soc of Expl Gophys, Expdd Abstrct Volum, (00), [9] Richrd Rochbrg d Kzuy Tchizw, Psudodiffrtil oprtors, Gbor frms, d locl trigoomtric bss, i Gbor lysis d lgorithms: Thory d pplictios, Hs G Fichtigr d Thoms Strohmr, ditors, Applid d Numricl Hrmoic Alysis, Birkhäusr, Bosto, 998 [0] Lurt Schwrtz, Théori ds oyux, Procdigs of th Itrtiol Cogrss of Mthmticis (950), CREWES Rsrch Rport Volum 5 (003)

Linear Algebra Existence of the determinant. Expansion according to a row.

Linear Algebra Existence of the determinant. Expansion according to a row. Lir Algbr 2270 1 Existc of th dtrmit. Expsio ccordig to row. W dfi th dtrmit for 1 1 mtrics s dt([]) = (1) It is sy chck tht it stisfis D1)-D3). For y othr w dfi th dtrmit s follows. Assumig th dtrmit

More information

Chapter 3 Fourier Series Representation of Periodic Signals

Chapter 3 Fourier Series Representation of Periodic Signals Chptr Fourir Sris Rprsttio of Priodic Sigls If ritrry sigl x(t or x[] is xprssd s lir comitio of som sic sigls th rspos of LI systm coms th sum of th idividul rsposs of thos sic sigls Such sic sigl must:

More information

page 11 equation (1.2-10c), break the bar over the right side in the middle

page 11 equation (1.2-10c), break the bar over the right side in the middle I. Corrctios Lst Updtd: Ju 00 Complx Vrils with Applictios, 3 rd ditio, A. Dvid Wusch First Pritig. A ook ought for My 007 will proly first pritig With Thks to Christi Hos of Swd pg qutio (.-0c), rk th

More information

COLLECTION OF SUPPLEMENTARY PROBLEMS CALCULUS II

COLLECTION OF SUPPLEMENTARY PROBLEMS CALCULUS II COLLECTION OF SUPPLEMENTARY PROBLEMS I. CHAPTER 6 --- Trscdtl Fuctios CALCULUS II A. FROM CALCULUS BY J. STEWART:. ( How is th umbr dfid? ( Wht is pproimt vlu for? (c ) Sktch th grph of th turl potil fuctios.

More information

National Quali cations

National Quali cations PRINT COPY OF BRAILLE Ntiol Quli ctios AH08 X747/77/ Mthmtics THURSDAY, MAY INSTRUCTIONS TO CANDIDATES Cdidts should tr thir surm, form(s), dt of birth, Scottish cdidt umbr d th m d Lvl of th subjct t

More information

Integration by Guessing

Integration by Guessing Itgrtio y Gussig Th computtios i two stdrd itgrtio tchiqus, Sustitutio d Itgrtio y Prts, c strmlid y th Itgrtio y Gussig pproch. This mthod cosists of thr stps: Guss, Diffrtit to chck th guss, d th Adjust

More information

Quantum Mechanics & Spectroscopy Prof. Jason Goodpaster. Problem Set #2 ANSWER KEY (5 questions, 10 points)

Quantum Mechanics & Spectroscopy Prof. Jason Goodpaster. Problem Set #2 ANSWER KEY (5 questions, 10 points) Chm 5 Problm St # ANSWER KEY 5 qustios, poits Qutum Mchics & Spctroscopy Prof. Jso Goodpstr Du ridy, b. 6 S th lst pgs for possibly usful costts, qutios d itgrls. Ths will lso b icludd o our futur ms..

More information

Some Common Fixed Point Theorems for a Pair of Non expansive Mappings in Generalized Exponential Convex Metric Space

Some Common Fixed Point Theorems for a Pair of Non expansive Mappings in Generalized Exponential Convex Metric Space Mish Kumr Mishr D.B.OhU Ktoch It. J. Comp. Tch. Appl. Vol ( 33-37 Som Commo Fi Poit Thorms for Pir of No psiv Mppigs i Grliz Epotil Cov Mtric Spc D.B.Oh Mish Kumr Mishr U Ktoch (Rsrch scholr Drvii Uivrsit

More information

[ ] Review. For a discrete-time periodic signal xn with period N, the Fourier series representation is

[ ] Review. For a discrete-time periodic signal xn with period N, the Fourier series representation is Discrt-tim ourir Trsform Rviw or discrt-tim priodic sigl x with priod, th ourir sris rprsttio is x + < > < > x, Rviw or discrt-tim LTI systm with priodic iput sigl, y H ( ) < > < > x H r rfrrd to s th

More information

National Quali cations

National Quali cations Ntiol Quli ctios AH07 X77/77/ Mthmtics FRIDAY, 5 MAY 9:00 AM :00 NOON Totl mrks 00 Attmpt ALL qustios. You my us clcultor. Full crdit will b giv oly to solutios which coti pproprit workig. Stt th uits

More information

SOLVED EXAMPLES. Ex.1 If f(x) = , then. is equal to- Ex.5. f(x) equals - (A) 2 (B) 1/2 (C) 0 (D) 1 (A) 1 (B) 2. (D) Does not exist = [2(1 h)+1]= 3

SOLVED EXAMPLES. Ex.1 If f(x) = , then. is equal to- Ex.5. f(x) equals - (A) 2 (B) 1/2 (C) 0 (D) 1 (A) 1 (B) 2. (D) Does not exist = [2(1 h)+1]= 3 SOLVED EXAMPLES E. If f() E.,,, th f() f() h h LHL RHL, so / / Lim f() quls - (D) Dos ot ist [( h)+] [(+h) + ] f(). LHL E. RHL h h h / h / h / h / h / h / h As.[C] (D) Dos ot ist LHL RHL, so giv it dos

More information

IX. Ordinary Differential Equations

IX. Ordinary Differential Equations IX. Orir Diffrtil Equtios A iffrtil qutio is qutio tht iclus t lst o rivtiv of uow fuctio. Ths qutios m iclu th uow fuctio s wll s ow fuctios of th sm vribl. Th rivtiv m b of orr thr m b svrl rivtivs prst.

More information

Emil Olteanu-The plane rotation operator as a matrix function THE PLANE ROTATION OPERATOR AS A MATRIX FUNCTION. by Emil Olteanu

Emil Olteanu-The plane rotation operator as a matrix function THE PLANE ROTATION OPERATOR AS A MATRIX FUNCTION. by Emil Olteanu Emil Oltu-Th pl rottio oprtor s mtri fuctio THE PLNE ROTTON OPERTOR S MTRX UNTON b Emil Oltu bstrct ormlism i mthmtics c offr m simplifictios, but it is istrumt which should b crfull trtd s it c sil crt

More information

Q.28 Q.29 Q.30. Q.31 Evaluate: ( log x ) Q.32 Evaluate: ( ) Q.33. Q.34 Evaluate: Q.35 Q.36 Q.37 Q.38 Q.39 Q.40 Q.41 Q.42. Q.43 Evaluate : ( x 2) Q.

Q.28 Q.29 Q.30. Q.31 Evaluate: ( log x ) Q.32 Evaluate: ( ) Q.33. Q.34 Evaluate: Q.35 Q.36 Q.37 Q.38 Q.39 Q.40 Q.41 Q.42. Q.43 Evaluate : ( x 2) Q. LASS XII Q Evlut : Q sc Evlut c Q Evlut: ( ) Q Evlut: Q5 α Evlut: α Q Evlut: Q7 Evlut: { t (t sc )} / Q8 Evlut : ( )( ) Q9 Evlut: Q0 Evlut: Q Evlut : ( ) ( ) Q Evlut : / ( ) Q Evlut: / ( ) Q Evlut : )

More information

Chapter 9 Infinite Series

Chapter 9 Infinite Series Sctio 9. 77. Cotiud d + d + C Ar lim b lim b b b + b b lim + b b lim + b b 6. () d (b) lim b b d (c) Not tht d c b foud by prts: d ( ) ( ) d + C. b Ar b b lim d lim b b b b lim ( b + ). b dy 7. () π dy

More information

MAT 182: Calculus II Test on Chapter 9: Sequences and Infinite Series Take-Home Portion Solutions

MAT 182: Calculus II Test on Chapter 9: Sequences and Infinite Series Take-Home Portion Solutions MAT 8: Clculus II Tst o Chptr 9: qucs d Ifiit ris T-Hom Portio olutios. l l l l 0 0 L'Hôpitl's Rul 0 . Bgi by computig svrl prtil sums to dvlop pttr: 6 7 8 7 6 6 9 9 99 99 Th squc of prtil sums is s follows:,,,,,

More information

ASSERTION AND REASON

ASSERTION AND REASON ASSERTION AND REASON Som qustios (Assrtio Rso typ) r giv low. Ech qustio cotis Sttmt (Assrtio) d Sttmt (Rso). Ech qustio hs choics (A), (B), (C) d (D) out of which ONLY ONE is corrct. So slct th corrct

More information

CLASS XI CHAPTER 3. Theorem 1 (sine formula) In any triangle, sides are proportional to the sines of the opposite angles. That is, in a triangle ABC

CLASS XI CHAPTER 3. Theorem 1 (sine formula) In any triangle, sides are proportional to the sines of the opposite angles. That is, in a triangle ABC CLSS XI ur I CHPTER.6. Proofs d Simpl pplictios of si d cosi formul Lt C b trigl. y gl w m t gl btw t sids d C wic lis btw 0 d 80. T gls d C r similrly dfid. T sids, C d C opposit to t vrtics C, d will

More information

LE230: Numerical Technique In Electrical Engineering

LE230: Numerical Technique In Electrical Engineering LE30: Numricl Tchiqu I Elctricl Egirig Lctur : Itroductio to Numricl Mthods Wht r umricl mthods d why do w d thm? Cours outli. Numbr Rprsttio Flotig poit umbr Errors i umricl lysis Tylor Thorm My dvic

More information

IV. The z-transform and realization of digital filters

IV. The z-transform and realization of digital filters www.tuworld.com www.tuworld.com Digitl Sigl Procssig 4 Dcmbr 6, 9 IV. Th -trsform d rlitio of digitl filtrs 7 Syllbus: Rviw of -trsforms, Applictios of -trsforms, Solutio of diffrc qutios of digitl filtrs,

More information

TOPIC 5: INTEGRATION

TOPIC 5: INTEGRATION TOPIC 5: INTEGRATION. Th indfinit intgrl In mny rspcts, th oprtion of intgrtion tht w r studying hr is th invrs oprtion of drivtion. Dfinition.. Th function F is n ntidrivtiv (or primitiv) of th function

More information

ENGI 3424 Appendix Formulæ Page A-01

ENGI 3424 Appendix Formulæ Page A-01 ENGI 344 Appdix Formulæ g A-0 ENGI 344 Egirig Mthmtics ossibilitis or your Formul Shts You my slct itms rom this documt or plcmt o your ormul shts. Howvr, dsigig your ow ormul sht c b vlubl rvisio xrcis

More information

PURE MATHEMATICS A-LEVEL PAPER 1

PURE MATHEMATICS A-LEVEL PAPER 1 -AL P MATH PAPER HONG KONG EXAMINATIONS AUTHORITY HONG KONG ADVANCED LEVEL EXAMINATION PURE MATHEMATICS A-LEVEL PAPER 8 am am ( hours) This papr must b aswrd i Eglish This papr cosists of Sctio A ad Sctio

More information

How much air is required by the people in this lecture theatre during this lecture?

How much air is required by the people in this lecture theatre during this lecture? 3 NTEGRATON tgrtio is us to swr qustios rltig to Ar Volum Totl qutity such s: Wht is th wig r of Boig 747? How much will this yr projct cost? How much wtr os this rsrvoir hol? How much ir is rquir y th

More information

Math 1272 Solutions for Fall 2004 Final Exam

Math 1272 Solutions for Fall 2004 Final Exam Mth 272 Solutios for Fll 2004 Fil Exm ) This itgrl pprs i Prolm of th udtd-2002? xm; solutio c foud i tht solutio st (B) 2) O of th first thigs tht should istigtd i plig th itgrtio of rtiol fuctio of polyomils

More information

Lectures 5-8: Fourier Series

Lectures 5-8: Fourier Series cturs 5-8: Fourir Sris PHY6 Rfrcs Jord & Smith Ch.6, Bos Ch.7, Kryszig Ch. Som fu jv pplt dmostrtios r vilbl o th wb. Try puttig Fourir sris pplt ito Googl d lookig t th sits from jhu, Flstd d Mths Oli

More information

terms of discrete sequences can only take values that are discrete as opposed to

terms of discrete sequences can only take values that are discrete as opposed to Diol Bgyoko () OWER SERIES Diitio Sris lik ( ) r th sm o th trms o discrt sqc. Th trms o discrt sqcs c oly tk vls tht r discrt s opposd to cotios, i.., trms tht r sch tht th mric vls o two cosctivs os

More information

Lecture 11 Waves in Periodic Potentials Today: Questions you should be able to address after today s lecture:

Lecture 11 Waves in Periodic Potentials Today: Questions you should be able to address after today s lecture: Lctur 11 Wvs in Priodic Potntils Tody: 1. Invrs lttic dfinition in 1D.. rphicl rprsnttion of priodic nd -priodic functions using th -xis nd invrs lttic vctors. 3. Sris solutions to th priodic potntil Hmiltonin

More information

z 1+ 3 z = Π n =1 z f() z = n e - z = ( 1-z) e z e n z z 1- n = ( 1-z/2) 1+ 2n z e 2n e n -1 ( 1-z )/2 e 2n-1 1-2n -1 1 () z

z 1+ 3 z = Π n =1 z f() z = n e - z = ( 1-z) e z e n z z 1- n = ( 1-z/2) 1+ 2n z e 2n e n -1 ( 1-z )/2 e 2n-1 1-2n -1 1 () z Sris Expasio of Rciprocal of Gamma Fuctio. Fuctios with Itgrs as Roots Fuctio f with gativ itgrs as roots ca b dscribd as follows. f() Howvr, this ifiit product divrgs. That is, such a fuctio caot xist

More information

ERDOS-SMARANDACHE NUMBERS. Sabin Tabirca* Tatiana Tabirca**

ERDOS-SMARANDACHE NUMBERS. Sabin Tabirca* Tatiana Tabirca** ERDO-MARANDACHE NUMBER b Tbrc* Tt Tbrc** *Trslv Uvrsty of Brsov, Computr cc Dprtmt **Uvrsty of Mchstr, Computr cc Dprtmt Th strtg pot of ths rtcl s rprstd by rct work of Fch []. Bsd o two symptotc rsults

More information

DTFT Properties. Example - Determine the DTFT Y ( e ) of n. Let. We can therefore write. From Table 3.1, the DTFT of x[n] is given by 1

DTFT Properties. Example - Determine the DTFT Y ( e ) of n. Let. We can therefore write. From Table 3.1, the DTFT of x[n] is given by 1 DTFT Proprtis Exampl - Dtrmi th DTFT Y of y α µ, α < Lt x α µ, α < W ca thrfor writ y x x From Tabl 3., th DTFT of x is giv by ω X ω α ω Copyright, S. K. Mitra Copyright, S. K. Mitra DTFT Proprtis DTFT

More information

Chapter 8 Approximation Methods, Hueckel Theory

Chapter 8 Approximation Methods, Hueckel Theory Witr 3 Chm 356: Itroductory Qutum Mchics Chptr 8 Approimtio Mthods, ucl Thory... 8 Approimtio Mthods... 8 Th Lir Vritiol Pricipl... mpl Lir Vritios... 3 Chptr 8 Approimtio Mthods, ucl Thory Approimtio

More information

Lectures 2 & 3 - Population ecology mathematics refresher

Lectures 2 & 3 - Population ecology mathematics refresher Lcturs & - Poultio cology mthmtics rrshr To s th mov ito vloig oultio mols, th olloig mthmtics crisht is suli I i out r mthmtics ttook! Eots logrithms i i q q q q q q ( tims) / c c c c ) ( ) ( Clculus

More information

1985 AP Calculus BC: Section I

1985 AP Calculus BC: Section I 985 AP Calculus BC: Sctio I 9 Miuts No Calculator Nots: () I this amiatio, l dots th atural logarithm of (that is, logarithm to th bas ). () Ulss othrwis spcifid, th domai of a fuctio f is assumd to b

More information

Discrete Fourier Transform (DFT)

Discrete Fourier Transform (DFT) Discrt Fourir Trasorm DFT Major: All Egirig Majors Authors: Duc guy http://umricalmthods.g.us.du umrical Mthods or STEM udrgraduats 8/3/29 http://umricalmthods.g.us.du Discrt Fourir Trasorm Rcalld th xpotial

More information

Chapter 5. Chapter 5 125

Chapter 5. Chapter 5 125 Chptr 5 Chptr 5: Itroductio to Digitl Filtrs... 6 5. Itroductio... 6 5.. No rcursiv digitl filtrs FIR... 7 5.. Rcursiv digitl filtr IIR... 8 5. Digitl Filtr Rlistio... 5.. Prlll rlistio... 5.. Cscd rlistio...

More information

TIME EVOLUTION OF SU(1,1) COHERENT STATES

TIME EVOLUTION OF SU(1,1) COHERENT STATES TIME EVOLUTION OF SU() COHERENT STATES J ZALEŚNY Istitut of Physics Tchicl Uivrsity of Szczci Al Pistów 48 7-3 Szczci Pold Mthmticl spcts of th SU() group prmtr ξ dymics govrd by Hmiltois xhibitig som

More information

A Single-Integral Representation for the Green Function of Steady Ship Flow in Water of Finite Depth

A Single-Integral Representation for the Green Function of Steady Ship Flow in Water of Finite Depth A Sigl-Itgrl Rprsttio for th Gr uctio of Stdy Ship low i Wtr of iit Dpth Thi Nguy & Xio-Bo Ch Costl Systs Sttio, P City, L 7 (USA x: --85-- Eil: NguyTC@csc.vy.il Rsrch Dprtt, BV, 7bis, Plc ds Rflts, 9

More information

Last time: introduced our first computational model the DFA.

Last time: introduced our first computational model the DFA. Lctur 7 Homwork #7: 2.2.1, 2.2.2, 2.2.3 (hnd in c nd d), Misc: Givn: M, NFA Prov: (q,xy) * (p,y) iff (q,x) * (p,) (follow proof don in clss tody) Lst tim: introducd our first computtionl modl th DFA. Tody

More information

Errata for Second Edition, First Printing

Errata for Second Edition, First Printing Errt for Scond Edition, First Printing pg 68, lin 1: z=.67 should b z=.44 pg 1: Eqution (.63) should rd B( R) = x= R = θ ( x R) p( x) R 1 x= [1 G( x)] = θp( R) + ( θ R)[1 G( R)] pg 15, problm 6: dmnd of

More information

Restricted Factorial And A Remark On The Reduced Residue Classes

Restricted Factorial And A Remark On The Reduced Residue Classes Applid Mathmatics E-Nots, 162016, 244-250 c ISSN 1607-2510 Availabl fr at mirror sits of http://www.math.thu.du.tw/ am/ Rstrictd Factorial Ad A Rmark O Th Rducd Rsidu Classs Mhdi Hassai Rcivd 26 March

More information

Session : Plasmas in Equilibrium

Session : Plasmas in Equilibrium Sssio : Plasmas i Equilibrium Ioizatio ad Coductio i a High-prssur Plasma A ormal gas at T < 3000 K is a good lctrical isulator, bcaus thr ar almost o fr lctros i it. For prssurs > 0.1 atm, collisio amog

More information

10. Joint Moments and Joint Characteristic Functions

10. Joint Moments and Joint Characteristic Functions 0 Joit Momts ad Joit Charactristic Fctios Followig sctio 6 i this sctio w shall itrodc varios paramtrs to compactly rprst th iformatio cotaid i th joit pdf of two rvs Giv two rvs ad ad a fctio g x y dfi

More information

INTEGRALS. Chapter 7. d dx. 7.1 Overview Let d dx F (x) = f (x). Then, we write f ( x)

INTEGRALS. Chapter 7. d dx. 7.1 Overview Let d dx F (x) = f (x). Then, we write f ( x) Chptr 7 INTEGRALS 7. Ovrviw 7.. Lt d d F () f (). Thn, w writ f ( ) d F () + C. Ths intgrls r clld indfinit intgrls or gnrl intgrls, C is clld constnt of intgrtion. All ths intgrls diffr y constnt. 7..

More information

PURE MATHEMATICS RESEARCH ARTICLE. Subjects: Science; Mathematics & Statistics; Advanced Mathematics; Algebra

PURE MATHEMATICS RESEARCH ARTICLE. Subjects: Science; Mathematics & Statistics; Advanced Mathematics; Algebra PURE MATHEMATICS RESEARCH ARTICLE Nturl rhotrix A.O. Isr * Rcivd: 27 July 206 Accptd: 05 Octobr 206 First Publishd: Octobr 206 *Corrspodig uthor: A.O. Isr, Dprtmt of Mthmtics, Ambros Alli Uivrsity, Ekpom

More information

Digital Signal Processing, Fall 2006

Digital Signal Processing, Fall 2006 Digital Sigal Procssig, Fall 6 Lctur 9: Th Discrt Fourir Trasfor Zhg-Hua Ta Dpartt of Elctroic Systs Aalborg Uivrsity, Dar zt@o.aau.d Digital Sigal Procssig, I, Zhg-Hua Ta, 6 Cours at a glac MM Discrt-ti

More information

Robust Estimation for ARMA models

Robust Estimation for ARMA models Robust Estimtio for ARMA modls Nor Mulr, Dil Pñ y d Víctor J. Yohi z Uivrsidd Torcuto di Tll, Uivrsidd Crlos III d Mdrid d Uivrsidd d Buos Airs d CONICET. Novmbr 8, 007 Abstrct This ppr itroducs w clss

More information

Figure 2-18 Thevenin Equivalent Circuit of a Noisy Resistor

Figure 2-18 Thevenin Equivalent Circuit of a Noisy Resistor .8 NOISE.8. Th Nyquist Nois Thorm W ow wat to tur our atttio to ois. W will start with th basic dfiitio of ois as usd i radar thory ad th discuss ois figur. Th typ of ois of itrst i radar thory is trmd

More information

Statistics 3858 : Likelihood Ratio for Exponential Distribution

Statistics 3858 : Likelihood Ratio for Exponential Distribution Statistics 3858 : Liklihood Ratio for Expotial Distributio I ths two xampl th rjctio rjctio rgio is of th form {x : 2 log (Λ(x)) > c} for a appropriat costat c. For a siz α tst, usig Thorm 9.5A w obtai

More information

Errata for Second Edition, First Printing

Errata for Second Edition, First Printing Errt for Scond Edition, First Printing pg 68, lin 1: z=.67 should b z=.44 pg 71: Eqution (.3) should rd B( R) = θ R 1 x= [1 G( x)] pg 1: Eqution (.63) should rd B( R) = x= R = θ ( x R) p( x) R 1 x= [1

More information

Derivation of a Predictor of Combination #1 and the MSE for a Predictor of a Position in Two Stage Sampling with Response Error.

Derivation of a Predictor of Combination #1 and the MSE for a Predictor of a Position in Two Stage Sampling with Response Error. Drivatio of a Prdictor of Cobiatio # ad th SE for a Prdictor of a Positio i Two Stag Saplig with Rspos Error troductio Ed Stak W driv th prdictor ad its SE of a prdictor for a rado fuctio corrspodig to

More information

On the Energy Complexity of LDPC Decoder Circuits

On the Energy Complexity of LDPC Decoder Circuits O th Ergy Complxity of LDPC Dcodr Circuits Christophr Blk d Frk R. Kschischg Dprtmt of Elctricl & Computr Egirig Uivrsity of Toroto christophr.blk@mil.utoroto.c frk@comm.utoroto.c rxiv:50.07999v [cs.it]

More information

A Simple Proof that e is Irrational

A Simple Proof that e is Irrational Two of th most bautiful ad sigificat umbrs i mathmatics ar π ad. π (approximatly qual to 3.459) rprsts th ratio of th circumfrc of a circl to its diamtr. (approximatly qual to.788) is th bas of th atural

More information

Discrete Fourier Transform. Discrete Fourier Transform. Discrete Fourier Transform. Discrete Fourier Transform. Discrete Fourier Transform

Discrete Fourier Transform. Discrete Fourier Transform. Discrete Fourier Transform. Discrete Fourier Transform. Discrete Fourier Transform Discrt Fourir Trasform Dfiitio - T simplst rlatio btw a lt- squc x dfid for ω ad its DTFT X ( ) is ω obtaid by uiformly sampli X ( ) o t ω-axis btw ω < at ω From t dfiitio of t DTFT w tus av X X( ω ) ω

More information

k m The reason that his is very useful can be seen by examining the Taylor series expansion of some potential V(x) about a minimum point:

k m The reason that his is very useful can be seen by examining the Taylor series expansion of some potential V(x) about a minimum point: roic Oscilltor Pottil W r ow goig to stuy solutios to t TIS for vry usful ottil tt of t roic oscilltor. I clssicl cics tis is quivlt to t block srig robl or tt of t ulu (for sll oscilltios bot of wic r

More information

H2 Mathematics Arithmetic & Geometric Series ( )

H2 Mathematics Arithmetic & Geometric Series ( ) H Mathmatics Arithmtic & Gomtric Sris (08 09) Basic Mastry Qustios Arithmtic Progrssio ad Sris. Th rth trm of a squc is 4r 7. (i) Stat th first four trms ad th 0th trm. (ii) Show that th squc is a arithmtic

More information

. Determine these to one correct decimal accuracy using the bisection method: (a) 2. The following equations all have a root in the interval ( 0,1.

. Determine these to one correct decimal accuracy using the bisection method: (a) 2. The following equations all have a root in the interval ( 0,1. PROBLEMS Us grhic rrsttio to dtrmi th zros of th followig fuctios to o corrct dciml : ( 4 4si ; (b ; (c ( ; (d 4 8 ; ( ; (f ; (g t I ordr to obti grhicl solutio of f ( o th itrvl [,b], ty th followig sttmts

More information

Solution to 1223 The Evil Warden.

Solution to 1223 The Evil Warden. Solutio to 1 Th Evil Ward. This is o of thos vry rar PoWs (I caot thik of aothr cas) that o o solvd. About 10 of you submittd th basic approach, which givs a probability of 47%. I was shockd wh I foud

More information

Vtusolution.in FOURIER SERIES. Dr.A.T.Eswara Professor and Head Department of Mathematics P.E.S.College of Engineering Mandya

Vtusolution.in FOURIER SERIES. Dr.A.T.Eswara Professor and Head Department of Mathematics P.E.S.College of Engineering Mandya LECTURE NOTES OF ENGINEERING MATHEMATICS III Su Cod: MAT) Vtusoutio.i COURSE CONTENT ) Numric Aysis ) Fourir Sris ) Fourir Trsforms & Z-trsforms ) Prti Diffrti Equtios 5) Lir Agr 6) Ccuus of Vritios Tt

More information

Discrete Fourier Transform. Nuno Vasconcelos UCSD

Discrete Fourier Transform. Nuno Vasconcelos UCSD Discrt Fourir Trasform uo Vascoclos UCSD Liar Shift Ivariat (LSI) systms o of th most importat cocpts i liar systms thory is that of a LSI systm Dfiitio: a systm T that maps [ ito y[ is LSI if ad oly if

More information

Minimum Spanning Trees

Minimum Spanning Trees Minimum Spnning Trs Minimum Spnning Trs Problm A town hs st of houss nd st of rods A rod conncts nd only houss A rod conncting houss u nd v hs rpir cost w(u, v) Gol: Rpir nough (nd no mor) rods such tht:

More information

Triple Play: From De Morgan to Stirling To Euler to Maclaurin to Stirling

Triple Play: From De Morgan to Stirling To Euler to Maclaurin to Stirling Tripl Play: From D Morga to Stirlig To Eulr to Maclauri to Stirlig Augustus D Morga (186-1871) was a sigificat Victoria Mathmaticia who mad cotributios to Mathmatics History, Mathmatical Rcratios, Mathmatical

More information

lecture 16: Introduction to Least Squares Approximation

lecture 16: Introduction to Least Squares Approximation 97 lecture 16: Itroductio to Lest Squres Approximtio.4 Lest squres pproximtio The miimx criterio is ituitive objective for pproximtig fuctio. However, i my cses it is more ppelig (for both computtio d

More information

Ch 1.2: Solutions of Some Differential Equations

Ch 1.2: Solutions of Some Differential Equations Ch 1.2: Solutions of Som Diffrntil Equtions Rcll th fr fll nd owl/mic diffrntil qutions: v 9.8.2v, p.5 p 45 Ths qutions hv th gnrl form y' = y - b W cn us mthods of clculus to solv diffrntil qutions of

More information

The Propagation Series

The Propagation Series //009 Th Progtio Sris rst /0 Th Progtio Sris Q: You rlir sttd tht sigl flow grhs r hlful i (cout m ) thr wys. I ow udrstd th first wy: Wy - Sigl flow grhs rovid us with grhicl ms of solvig lrg systms of

More information

Walk Like a Mathematician Learning Task:

Walk Like a Mathematician Learning Task: Gori Dprtmnt of Euction Wlk Lik Mthmticin Lrnin Tsk: Mtrics llow us to prform mny usful mthmticl tsks which orinrily rquir lr numbr of computtions. Som typs of problms which cn b on fficintly with mtrics

More information

ME 522 PRINCIPLES OF ROBOTICS. FIRST MIDTERM EXAMINATION April 19, M. Kemal Özgören

ME 522 PRINCIPLES OF ROBOTICS. FIRST MIDTERM EXAMINATION April 19, M. Kemal Özgören ME 522 PINCIPLES OF OBOTICS FIST MIDTEM EXAMINATION April 9, 202 Nm Lst Nm M. Kml Özgörn 2 4 60 40 40 0 80 250 USEFUL FOMULAS cos( ) cos cos sin sin sin( ) sin cos cos sin sin y/ r, cos x/ r, r 0 tn 2(

More information

The Propagation Series

The Propagation Series /9/009 Th Progtio Sris.doc /8 Th Progtio Sris Q: You rlir sttd tht sigl flow grhs r hlful i (cout m ) thr wys. I ow udrstd th first wy: Wy - Sigl flow grhs rovid us with grhicl ms of solvig lrg systms

More information

POWER SERIES R. E. SHOWALTER

POWER SERIES R. E. SHOWALTER POWER SERIES R. E. SHOWALTER. sequeces We deote by lim = tht the limit of the sequece { } is the umber. By this we me tht for y ε > 0 there is iteger N such tht < ε for ll itegers N. This mkes precise

More information

Advanced Engineering Mathematics, K.A. Stroud, Dexter J. Booth Engineering Mathematics, H.K. Dass Higher Engineering Mathematics, Dr. B.S.

Advanced Engineering Mathematics, K.A. Stroud, Dexter J. Booth Engineering Mathematics, H.K. Dass Higher Engineering Mathematics, Dr. B.S. Rfrc: (i) (ii) (iii) Advcd Egirig Mhmic, K.A. Sroud, Dxr J. Booh Egirig Mhmic, H.K. D Highr Egirig Mhmic, Dr. B.S. Grwl Th mhod of m Thi coi of h followig xm wih h giv coribuio o h ol. () Mid-rm xm : 3%

More information

Integration Continued. Integration by Parts Solving Definite Integrals: Area Under a Curve Improper Integrals

Integration Continued. Integration by Parts Solving Definite Integrals: Area Under a Curve Improper Integrals Intgrtion Continud Intgrtion y Prts Solving Dinit Intgrls: Ar Undr Curv Impropr Intgrls Intgrtion y Prts Prticulrly usul whn you r trying to tk th intgrl o som unction tht is th product o n lgric prssion

More information

Review of the Riemann Integral

Review of the Riemann Integral Chpter 1 Review of the Riem Itegrl This chpter provides quick review of the bsic properties of the Riem itegrl. 1.0 Itegrls d Riem Sums Defiitio 1.0.1. Let [, b] be fiite, closed itervl. A prtitio P of

More information

(HELD ON 22nd MAY SUNDAY 2016) MATHEMATICS CODE - 2 [PAPER -2]

(HELD ON 22nd MAY SUNDAY 2016) MATHEMATICS CODE - 2 [PAPER -2] QUESTION PAPER WITH SOLUTION OF JEE ADVANCED - 6 7. Lt P (HELD ON d MAY SUNDAY 6) FEEL THE POWER OF OUR KNOWLEDGE & EXPERIENCE Our Top clss IITi fculty tm promiss to giv you uthtic swr ky which will b

More information

Time : 1 hr. Test Paper 08 Date 04/01/15 Batch - R Marks : 120

Time : 1 hr. Test Paper 08 Date 04/01/15 Batch - R Marks : 120 Tim : hr. Tst Papr 8 D 4//5 Bch - R Marks : SINGLE CORRECT CHOICE TYPE [4, ]. If th compl umbr z sisfis th coditio z 3, th th last valu of z is qual to : z (A) 5/3 (B) 8/3 (C) /3 (D) o of ths 5 4. Th itgral,

More information

The Matrix Exponential

The Matrix Exponential Th Matrix Exponntial (with xrciss) by D. Klain Vrsion 207.0.05 Corrctions and commnts ar wlcom. Th Matrix Exponntial For ach n n complx matrix A, dfin th xponntial of A to b th matrix A A k I + A + k!

More information

TRANSFORMS AND PARTIAL DIFFERENTIAL EQUATIONS

TRANSFORMS AND PARTIAL DIFFERENTIAL EQUATIONS TRANSFORMS AND PARTIAL DIFFERENTIAL EQUATIONS UNIT-I PARTIAL DIFFERENTIAL EQUATIONS PART-A. Elimit th ritrry ott & from = ( + )(y + ) Awr: = ( + )(y + ) Diff prtilly w.r.to & y hr p & q y p = (y + ) ;

More information

We will begin by supplying the proof to (a).

We will begin by supplying the proof to (a). (The solutios of problem re mostly from Jeffrey Mudrock s HWs) Problem 1. There re three sttemet from Exmple 5.4 i the textbook for which we will supply proofs. The sttemets re the followig: () The spce

More information

1 Introduction to Modulo 7 Arithmetic

1 Introduction to Modulo 7 Arithmetic 1 Introution to Moulo 7 Arithmti Bor w try our hn t solvin som hr Moulr KnKns, lt s tk los look t on moulr rithmti, mo 7 rithmti. You ll s in this sminr tht rithmti moulo prim is quit irnt rom th ons w

More information

PREPARATORY MATHEMATICS FOR ENGINEERS

PREPARATORY MATHEMATICS FOR ENGINEERS CIVE 690 This qusti ppr csists f 6 pritd pgs, ch f which is idtifid by th Cd Numbr CIVE690 FORMULA SHEET ATTACHED UNIVERSITY OF LEEDS Jury 008 Emiti fr th dgr f BEg/ MEg Civil Egirig PREPARATORY MATHEMATICS

More information

On the approximation of the constant of Napier

On the approximation of the constant of Napier Stud. Uiv. Babş-Bolyai Math. 560, No., 609 64 O th approximatio of th costat of Napir Adri Vrscu Abstract. Startig from som oldr idas of [] ad [6], w show w facts cocrig th approximatio of th costat of

More information

Digital Signal Processing. Hossein Mahvash Mohammadi

Digital Signal Processing. Hossein Mahvash Mohammadi Digitl Sigl Procssig Hossi Mhvsh Mohmmdi h.mhvsh@g.ui.c.ir Rfrcs Discrt-Tim Sigl Procssig b Opphim A.V., Schfr R.W, d Editio, Prtic-Hll, 999. Digitl Sigl Procssig with Computr Applictios, P. A. L, W. Furst

More information

The Matrix Exponential

The Matrix Exponential Th Matrix Exponntial (with xrciss) by Dan Klain Vrsion 28928 Corrctions and commnts ar wlcom Th Matrix Exponntial For ach n n complx matrix A, dfin th xponntial of A to b th matrix () A A k I + A + k!

More information

Limits Indeterminate Forms and L Hospital s Rule

Limits Indeterminate Forms and L Hospital s Rule Limits Indtrmint Forms nd L Hospitl s Rul I Indtrmint Form o th Tp W hv prviousl studid its with th indtrmint orm s shown in th ollowin mpls: Empl : Empl : tn [Not: W us th ivn it ] Empl : 8 h 8 [Not:

More information

DETERMINANT. = 0. The expression a 1. is called a determinant of the second order, and is denoted by : y + c 1

DETERMINANT. = 0. The expression a 1. is called a determinant of the second order, and is denoted by : y + c 1 NOD6 (\Dt\04\Kot\J-Advced\SMP\Mths\Uit#0\NG\Prt-\0.Determits\0.Theory.p65. INTRODUCTION : If the equtios x + b 0, x + b 0 re stisfied by the sme vlue of x, the b b 0. The expressio b b is clled determit

More information

Infinite Series Sequences: terms nth term Listing Terms of a Sequence 2 n recursively defined n+1 Pattern Recognition for Sequences Ex:

Infinite Series Sequences: terms nth term Listing Terms of a Sequence 2 n recursively defined n+1 Pattern Recognition for Sequences Ex: Ifiite Series Sequeces: A sequece i defied s fuctio whose domi is the set of positive itegers. Usully it s esier to deote sequece i subscript form rther th fuctio ottio.,, 3, re the terms of the sequece

More information

Part B: Transform Methods. Professor E. Ambikairajah UNSW, Australia

Part B: Transform Methods. Professor E. Ambikairajah UNSW, Australia Part B: Trasform Mthods Chaptr 3: Discrt-Tim Fourir Trasform (DTFT) 3. Discrt Tim Fourir Trasform (DTFT) 3. Proprtis of DTFT 3.3 Discrt Fourir Trasform (DFT) 3.4 Paddig with Zros ad frqucy Rsolutio 3.5

More information

Chapter 2 Infinite Series Page 1 of 11. Chapter 2 : Infinite Series

Chapter 2 Infinite Series Page 1 of 11. Chapter 2 : Infinite Series Chatr Ifiit Sris Pag of Sctio F Itgral Tst Chatr : Ifiit Sris By th d of this sctio you will b abl to valuat imror itgrals tst a sris for covrgc by alyig th itgral tst aly th itgral tst to rov th -sris

More information

Problem Value Score Earned No/Wrong Rec -3 Total

Problem Value Score Earned No/Wrong Rec -3 Total GEORGIA INSTITUTE OF TECHNOLOGY SCHOOL of ELECTRICAL & COMPUTER ENGINEERING ECE6 Fall Quiz # Writt Eam Novmr, NAME: Solutio Kys GT Usram: LAST FIRST.g., gtiit Rcitatio Sctio: Circl t dat & tim w your Rcitatio

More information

TRANSFORMS AND PARTIAL DIFFERENTIAL EQUATIONS

TRANSFORMS AND PARTIAL DIFFERENTIAL EQUATIONS TRANSFORMS AND PARTIAL DIFFERENTIAL EQUATIONS UNIT-I PARTIAL DIFFERENTIAL EQUATIONS PART-A. Elimit th ritrry ott & from = ( + )(y + ) = ( + )(y + ) Diff prtilly w.r.to & y hr p & q p = (y + ) ; q = ( +

More information

Chapter Five. More Dimensions. is simply the set of all ordered n-tuples of real numbers x = ( x 1

Chapter Five. More Dimensions. is simply the set of all ordered n-tuples of real numbers x = ( x 1 Chatr Fiv Mor Dimsios 51 Th Sac R W ar ow rard to mov o to sacs of dimsio gratr tha thr Ths sacs ar a straightforward gralizatio of our Euclida sac of thr dimsios Lt b a ositiv itgr Th -dimsioal Euclida

More information

Thermodynamic Properties and XAFS Debye Waller Factors of Metallic Nickel

Thermodynamic Properties and XAFS Debye Waller Factors of Metallic Nickel Itrtiol Jourl of Modr Physics d Applictios Vol. No. 5 pp. -8 http:www.iscic.orgjourlijmp Thrmodymic Proprtis d XAFS Dby Wllr Fctors of Mtllic Nickl Nguy V Hug * Dih Quoc Vuog Dprtmt of Physics Collg of

More information

Chapter 16. 1) is a particular point on the graph of the function. 1. y, where x y 1

Chapter 16. 1) is a particular point on the graph of the function. 1. y, where x y 1 Prctic qustions W now tht th prmtr p is dirctl rltd to th mplitud; thrfor, w cn find tht p. cos d [ sin ] sin sin Not: Evn though ou might not now how to find th prmtr in prt, it is lws dvisl to procd

More information

On Gaussian Distribution

On Gaussian Distribution Prpr b Çğt C MTU ltril gi. Dpt. 30 Sprig 0089 oumt vrio. Gui itributio i i ollow O Gui Ditributio π Th utio i lrl poitiv vlu. Bor llig thi utio probbilit it utio w houl h whthr th r ur th urv i qul to

More information

Chapter 11.00C Physical Problem for Fast Fourier Transform Civil Engineering

Chapter 11.00C Physical Problem for Fast Fourier Transform Civil Engineering haptr. Physical Problm for Fast Fourir Trasform ivil Egirig Itroductio I this chaptr, applicatios of FFT algorithms [-5] for solvig ral-lif problms such as computig th dyamical (displacmt rspos [6-7] of

More information

COMPUTER GENERATED HOLOGRAMS Optical Sciences 627 W.J. Dallas (Monday, April 04, 2005, 8:35 AM) PART I: CHAPTER TWO COMB MATH.

COMPUTER GENERATED HOLOGRAMS Optical Sciences 627 W.J. Dallas (Monday, April 04, 2005, 8:35 AM) PART I: CHAPTER TWO COMB MATH. C:\Dallas\0_Courss\03A_OpSci_67\0 Cgh_Book\0_athmaticalPrliminaris\0_0 Combath.doc of 8 COPUTER GENERATED HOLOGRAS Optical Scincs 67 W.J. Dallas (onday, April 04, 005, 8:35 A) PART I: CHAPTER TWO COB ATH

More information

cycle that does not cross any edges (including its own), then it has at least

cycle that does not cross any edges (including its own), then it has at least W prov th following thorm: Thorm If a K n is drawn in th plan in such a way that it has a hamiltonian cycl that dos not cross any dgs (including its own, thn it has at last n ( 4 48 π + O(n crossings Th

More information

EEE 303: Signals and Linear Systems

EEE 303: Signals and Linear Systems 33: Sigls d Lir Sysms Orhogoliy bw wo sigls L us pproim fucio f () by fucio () ovr irvl : f ( ) = c( ); h rror i pproimio is, () = f() c () h rgy of rror sigl ovr h irvl [, ] is, { }{ } = f () c () d =

More information

NEW VERSION OF SZEGED INDEX AND ITS COMPUTATION FOR SOME NANOTUBES

NEW VERSION OF SZEGED INDEX AND ITS COMPUTATION FOR SOME NANOTUBES Digst Joural of Naomatrials ad Biostructurs Vol 4, No, March 009, p 67-76 NEW VERSION OF SZEGED INDEX AND ITS COMPUTATION FOR SOME NANOTUBES A IRANMANESH a*, O KHORMALI b, I NAJAFI KHALILSARAEE c, B SOLEIMANI

More information

( a n ) converges or diverges.

( a n ) converges or diverges. Chpter Ifiite Series Pge of Sectio E Rtio Test Chpter : Ifiite Series By the ed of this sectio you will be ble to uderstd the proof of the rtio test test series for covergece by pplyig the rtio test pprecite

More information

CAPACITANCE CALCULATION USING EEM

CAPACITANCE CALCULATION USING EEM Act Elctrotchic t Iformtic Vol. 9, o., 009, 5 3 5 CAPACITACE CALCULATIO USIG EEM Slvoljub R. ALEKSIĆ, Mirj PERIĆ, Sš S. ILIĆ Dprtmt of Thorticl Elctricl Egirig, Fculty of Elctroic Egirig, Uivrsity of iš,

More information