Joint source and relay design for MIMO multi-relay systems using projected gradient approach

Size: px
Start display at page:

Download "Joint source and relay design for MIMO multi-relay systems using projected gradient approach"

Transcription

1 Todng et al EURASIP Journal on Wreless Communcatons and Networkng 04, 04:5 RESEARC Open Access Jont source and relay desgn for MIMO mult-relay systems usng projected gradent approach Aprana Todng, Muhammad RA Khandaker and Yue Rong 3* Abstract In ths paper, we develop the optmal source precodng matrx and relay amplfyng matrces for non-regeneratve multple-nput multple-output MIMO relay communcaton systems wth parallel relay nodes usng the projected gradent PG approach We show that the optmal relay amplfyng matrces have a beamformng structure Explotng the structure of relay matrces, an teratve jont source and relay matrces optmzaton algorthm s developed to mnmze the mean-squared error MSE of the sgnal waveform estmaton at the destnaton usng the PG approach The performance of the proposed algorthm s demonstrated through numercal smulatons Keywords: MIMO relay; Parallel relay network; Beamformng; Non-regeneratve relay; Projected gradent Introducton Recently, multple-nput multple-output MIMO relay communcaton systems have attracted much research nterest and provded sgnfcant mprovement n terms of both spectral effcency and lnk relablty [-7] Many works have studed the optmal relay amplfyng matrx for the source-relay-destnaton channel In [,3], the optmal relay amplfyng matrx maxmzng the mutual nformaton MI between the source and destnaton nodes was derved, assumng that the source covarance matrx s an dentty matrx In [4-6], the optmal relay amplfyng matrx was desgned to mnmze the mean-squared error MSE of the sgnal waveform estmaton at the destnaton A few research has studed the jont optmzaton of the source precodng matrx and the relay amplfyng matrx for the source-relay-destnaton channel In [7], both the source and relay matrces were jontly desgned to maxmze the source-destnaton MI In [8,9], source and relay matrces were developed to jontly optmze a broad class of objectve functons The author of [0] nvestgated the *Correspondence: yrong@curtneduau 3 Department of Electrcal and Computer Engneerng, Curtn Unversty of Technology, Bentley, WA 60, Australa Full lst of author nformaton s avalable at the end of the artcle jont source and relay optmzaton for two-way MIMO relay systems usng the projected gradent PG approach The source and relay optmzaton for mult-user MIMO relay systems wth sngle relay node has been nvestgated n [-4] All the works n [-4] consdered a sngle relay node at each hop In general, jont source and relay precodng matrces desgn for MIMO relay systems wth multple relay nodes s more challengng than that for sngle-relay systems The authors of [5] developed the optmal relay amplfyng matrces wth multple relay nodes A matrxform conjugate gradent algorthm has been proposed n [6] to optmze the source and relay matrces In [7], the authors proposed a suboptmal source and relay matrces desgn for parallel MIMO relay systems by frst relaxng the power constrant at each relay node to the sum relay power constrants at the output of the second-hop channel and then scalng the relay matrces to satsfy the ndvdual relay power constrants In ths paper, we propose a jontly optmal source precodng matrx and relay amplfyng matrces desgn for a two-hop non-regeneratve MIMO relay network wth multple relay nodes usng the PG approach We show that the optmal relay amplfyng matrces have a beamformng structure Ths new result s not avalable n [6] It generalzes the optmal source and relay matrces desgn 04 Todng et al; lcensee Sprnger Ths s an Open Access artcle dstrbuted under the terms of the Creatve Commons Attrbuton Lcense whch permts unrestrcted use, dstrbuton, and reproducton n any medum, provded the orgnal work s properly credted

2 Todng et al EURASIP Journal on Wreless Communcatons and Networkng 04, 04:5 Page of 9 from a sngle relay node case [8] to multple parallel relay nodes scenaros Explotng the structure of relay matrces, an teratve jont source and relay matrces optmzaton algorthm s developed to mnmze the MSE of the sgnal waveform estmaton Dfferent to [7], n ths paper, we develop the optmal source and relay matrces by drectly consderng the transmsson power constrant at each relay node Smulaton results demonstrate the effectveness of the proposed teratve jont source and relay matrces desgn algorthm wth multple parallel relay nodes usng the PG approach The rest of ths paper s organzed as follows In Secton, we ntroduce the model of a non-regeneratve MIMO relay communcaton system wth parallel relay nodes The jont source and relay matrces desgn algorthm s developed n Secton 3 In Secton 4, we show some numercal smulatons Conclusons are drawn n Secton 5 Systemmodel In ths secton, we ntroduce the model of a two-hop MIMO relay communcaton system consstng of one source node, K parallel relay nodes, and one destnaton node as shown n Fgure We assume that the source and destnaton nodes have N s and N d antennas, respectvely, and each relay node has N r antennas The generalzaton to systems wth dfferent number of antennas at each relay node s straghtforward Due to ts mert of smplcty, a lnear non-regeneratve strategy s appled at each relay node The communcaton process between the source and destnaton nodes s completed n two tme slots In the frst tme slot, the N b N b N s modulated source symbol vector s s lnearly precoded as x = Bs, where B s an N s N b source precodng matrx We assume that the source sgnal vector satsfes E [ ss ] = I Nb,whereI n stands for an n n dentty matrx, s the matrx vector ermtan transpose, and E[ ] denotes statstcal expectaton The precoded vector x s transmtted to K parallel relay nodes The N r recevedsgnal vector at the th relay node can be wrtten as y r, = sr, x + v r,, =,, K, where sr, s the N r N s MIMO channel matrx between thesourceandtheth relay nodes and v r, s the addtve Gaussannosevectoratthe th relay node Inthesecondtmeslot,thesourcenodesslent,whle each relay node transmts the lnearly amplfed sgnal vector to the destnaton node as x r, = F y r,, =,, K, 3 where F s the N r N r amplfyng matrx at the th relay node The receved sgnal vector at the destnaton node can be wrtten as y d = rd, x r, + v d, 4 = where rd, s the N d N r MIMO channel matrx between the th relay and the destnaton nodes, and v d s the addtve Gaussan nose vector at the destnaton node v r v d F x N r N r y s B x x Ns sr, N r F N r rd, y W y Nd ŝ F K N r N r Source Relay Destnaton Fgure Block dagram of a parallel MIMO relay communcaton system

3 Todng et al EURASIP Journal on Wreless Communcatons and Networkng 04, 04:5 Page 3 of 9 Substtutng to 3 nto 4, we have y d = rd, F sr, Bs + rd, F v r, + v d = = rd F sr Bs + rd Fv r + v d s +ṽ, 5 [ ] T where sr T sr,, T sr,,, T sr,k s a KNr N s channel matrx between the source node and all relay nodes, rd [ ] rd,, rd,,, rd,k s an Nd KN r channel matrx between all relay nodes and the destnaton node, F bd[f, F,, F K ]sthekn r KN r block dagonal equvalent relay matrx, v r [ v T r,, vt r,,, ] T vt r,k s obtaned by stackng the nose vectors at all the relays, rd F sr B s the effectve MIMO channel matrx of the source-relay-destnaton lnk, and ṽ rd Fv r + v d s the equvalent nose vector ere, T denotes the matrx vector transpose, bd[ ] constructsablock- dagonal matrx We assume that all noses are ndependent and dentcally dstrbuted d Gaussan nose wth zero mean and unt varance The transmsson power consumed by each relay node 3 can be expressed as E [ tr x r, x [ r,] = tr F sr, BB sr, + I ] N r F, =,, K, 6 where tr stands for the matrx trace Usng a lnear recever, the estmated sgnal waveform vector at the destnaton node s gven by ŝ = W y d, where W s an N d N b weght matrx The MSE of the sgnal waveform estmaton s gven by [ ŝ ŝ ] MSE = tr E s s = tr W I Nb W I Nb + W CW, 7 where C s the equvalent nose covarance matrx gven by C = E [ ṽṽ ] = rd FF rd + I N d The weght matrx W whch mnmzes 7 s the Wener flter and can be wrtten as W = + C, 8 where denotes the matrx nverson Substtutng 8 back nto 7, t can be seen that the MSE s a functon of F and B and can be wrtten as [ ] MSE = tr I + Nb C 9 3 Jont source and relay matrx optmzaton In ths secton, we address the jont source and relay matrx optmzaton problem for MIMO mult-relay systems wth a lnear mnmum mean-squared error MMSE recever at the destnaton node In partcular, we show that optmal relay matrces have a general beamformng structure Based on 6 and 9, the jont source and relay matrces optmzaton problem can be formulated as [ mn tr I + Nb C ] 0 {F },B st tr BB P s tr [ F sr, BB sr, + I ] N r F P r,, =,, K, where {F } {F, =,, K}, s the transmt power constrant at the source node, whle s the power constrant at each relay node ere, P s > 0and P r, > 0, =,, K, are the correspondng power budget Obvously, to avod any loss of transmsson power n the relay system when a lnear recever s used, there should be N b mnkn r, N d The problem 0- s non-convex, and a globally optmal soluton of B and {F } s dffcult to obtan wth a reasonable computatonal complexty In ths paper, we develop an teratve algorthm to optmze B and {F } Frst, we show the optmal structure of {F } 3 Optmal structure of relay amplfyng matrces For gven source matrx B satsfyng, the relay matrces {F } are optmzed by solvng the followng problem: mn {F } [ ] tr I + Nb C 3 st tr F [ sr, BB sr, + I N r ] F Pr,, =,, K 4 Let us ntroduce the followng sngular value decompostons SVDs: sr, B = U s, s, V s,, rd, = U r, r, V r,, =,, K, 5 where s, and r, are R s, R s, and R r, R r, dagonal matrces, respectvely ere, R s, rank sr, B, R r, rank rd,, =,, K, and rank denotes the rank of a matrx Based on the defnton of matrx rank, R s, mnn r, N b and R r, mnn r, N d The followng theorem states the structure of the optmal {F } Theorem Usng the SVDs of 5, the optmal structure of F as the soluton to the problem 3-4 s gven by F = V r, A U s,, =,, K, 6 where A s an R r, R s, matrx, =,, K Proof See Appendx

4 Todng et al EURASIP Journal on Wreless Communcatons and Networkng 04, 04:5 Page 4 of 9 The remanng task s to fnd the optmal A, =,, K From 3 and 3 n Appendx, we can equvalently rewrte the optmzaton problem 3-4 as [ mn tr I Nb + V s, s, A r, U r, {A } = K U r, r, A A r, U r, + I N d = ] U r, r, A s, V s, 7 = st tr A s, + I R s, A Pr,, =,, K 8 Both the problem 3-4 and the problem 7-8 have matrx optmzaton varables owever, n the former problem, the optmzaton varable F s an N r N r matrx, whle the dmenson of A s R r, R s,, whch may be smaller than that of F Thus, solvng the problem 7-8 has a smaller computatonal complexty than solvng the problem 3-4 In general, the problem 7-8 s non-convex, and a globally optmal soluton s dffcult to obtan wth a reasonable computatonal complexty Fortunately, we can resort to numercal methods, such as the projected gradent algorthm [8] to fnd at least a locally optmal soluton of 7-8 Theorem Let us defne the objectve functon n 7 as f A Its gradent f A wth respect to A can be calculated by usng results on dervatves of matrces n [9] as E S + D R G E S, =,, K, 9 f A = R M where M, R, S, D, E,andG are defned n Appendx Proof See Appendx In each teraton of the PG algorthm, we frst obtan à = A s n f A by movng A one step towards the negatve gradent drecton of f A,wheres n > 0sthe step sze Snce à mght not satsfy the constrant 8, we need to project t onto the set gven by 8 The projected matrx Ā s obtaned by mnmzng the Frobenus norm of Ā à accordng to [8] subjectng to 8, whch can be formulated as the followng optmzaton problem: mn Ā tr Ā à Ā à 0 st tr Ā s, + I Rs, Ā Pr, Obvously, f trã s, + I Rs, à P r,,thenā = à Otherwse, the soluton to the problem 0- can be obtaned by usng the Lagrange multpler method, and the soluton s gven by [ Ā = à λ + IRs, + λ s,], where λ>0sthe soluton to the non-lnear equaton of [ tr à λ + IRs, + s,] λ s, + I Rs, [ λ + I Rs, + λ s,] à = P r, Equaton can be effcently solved by the bsecton method [8] The procedure of the PG algorthm s lsted n Algorthm, where n denotes the varable at the nth teraton, δ n and s n are the step sze parameters at the nth teraton, denotes the maxmum among the absolute value of all elements n the matrx, and ε s a postve constant close to 0 The step sze parameters δ n and s n are determned by the Armjo rule [8], e, s n = s s a constant through all teratons, whle at the nth teraton, δ n s set to be γ m n ere, m n s the mnmal non-negatve nteger that satsfes the followng nequalty f A n+ f A n αγ ntr m Ān f,whereα and γ are A n A n constants Accordng to [8], usually α s chosen close to 0, for example, α [0 5,0 ], whle a proper choce of γ s normally from 0 to 05 3 Optmal source precodng matrx Wth fxed {F }, the source precodng matrx B s optmzed by solvng the followng problem: Algorthm Procedure of applyng the projected gradent algorthm to solve the problem 7-8 Intalze the algorthm at a feasble A 0 for =,, K;setn = 0 For =,, K, Compute the gradent of 7 f A n ; Project à n = A n s n f A n to obtan Ā n ; Update A wth A n+ = A n Ān + δ n A n 3 If max A n+ A n ε, then end Otherwse, let n := n + andgotostep mn B [INb tr + B B ] 3 st tr BB P s, 4 tr F sr, BB sr, F P r,, =,, K, 5

5 Todng et al EURASIP Journal on Wreless Communcatons and Networkng 04, 04:5 Page 5 of 9 where sr F rd rd FF rd + I N d rd F sr,and P r, P r, tr F F, =,, K Letusntroduce BB, and a postve sem-defnte PSD matrx X, wth X I Ns + wherea B means that A B s a PSD matrx By usng the Schur complement [0], the problem 3-5 can be equvalently converted to the followng problem: mn tr X N s + N b 6 X, X INs st I Ns I Ns + 0, 7 tr P s, 0, 8 tr F sr, sr, F P r,, =,, K 9 The problem 6-9 s a convex sem-defnte programmng SDP problem whch can be effcently solved by the nteror pont method [0] Let us ntroduce the egenvalue decomposton EVD of = U U, where s a R R egenvalue matrx wth R = rank IfR = N b,thenfrom = BB,wehave B = U IfR > N b, the randomzaton technque [] can be appled to obtan a possbly suboptmal soluton of B wth rank N b IfR < N b, t ndcates that the system channel cannot support N b ndependent data streams, and thus, n ths case, a smaller N b should be chosen n the system desgn Now, the orgnal jont source and relay optmzaton problem 0- can be solved by an teratve algorthm as shown n Algorthm, where m denotes the varable at the mth teraton Ths algorthm s frst ntalzed at a random feasble B satsfyng At each teraton, we frst update {F } wth fxed B and then update B wth fxed {F } Note that the condtonal updates of each matrx may ether decrease or mantan but cannot ncrease the objectve functon 0 Monotonc convergence of {F } and B towards at least a locally optmal soluton follows drectly from ths observaton Note that n each teraton of ths algorthm, we need to update the relay amplfyng matrces accordng to the procedure lsted n Algorthm at a complexty order of O K Nd 3 + N3 r + b N3 and update the source precodng matrx through solvng the SDP problem 6-9 at a N complexty cost that s at most O s + K + 35 usng nteror pont methods [] Therefore, the per-teraton computatonal complexty order of the proposed algorthm s O K Nd 3 + N3 r + b N3 + N s + K + 35 The overall complexty of ths algorthm depends on the number of teratons untl convergence, whch wll be studed n the next secton Algorthm Procedure of solvng the problem 0- Intalze the algorthm at a feasble B 0 satsfyng constrant ; set m = 0 For fxed B m, obtan {F } m by solvng the problem 7-8 usng the PG algorthm 3 Update B m+ by solvng the problem 6-9 wth known {F } m 4 If B m+ B m ε, then end Otherwse, let m := m + andgotostep 4 Smulatons In ths secton, we study the performance of the proposed jontly optmal source and relay matrx desgn for MIMO mult-relay systems wth lnear MMSE recever All smulatons are conducted n a flat Raylegh fadng envronment where the channel matrces have zero-mean entres wth varances σ s /N s and σ r /KN r for sr and rd, respectvely For the sake of smplcty, we assume P r, = P r, =,, K The BPSK constellatons are used to modulate the source symbols, and all noses are d Gaussan wth zero mean and unt varance We defne SNR s = σ s P skn r /N s and SNR r = σ r P rn d /KN r as the sgnal-to-nose rato SNR for the source-relay lnk and the relay-destnaton lnk, respectvely We transmt 000N s randomly generated bts n each channel realzaton, and all smulaton results are averaged over 00 channel realzatons In all smulatons, we set N b = N s = N r = N d = 3, and the MMSE lnear recever n 8 s employed at the destnaton for symbol detecton In the frst example, a MIMO relay system wth K = 3 relay nodes s smulated We compare the normalzed MSE performance of the proposed jont source and relay optmzaton algorthm usng the projected gradent JSR- PG algorthm n Algorthm, the optmal relay-only algorthm usng the projected gradent ORO-PG algorthm n Algorthm wth B = P s /N s I Ns,andthenave amplfy-and-forward NAF algorthm Fgure shows the normalzed MSE of all algorthms versus SNR s for SNR r = 0 db Whle Fgure 3 demonstrates the normalzed MSE of all algorthms versus SNR r for SNR s fxed at 0 db It can be seen from Fgures and 3 that the JSR- PG and ORO-PG algorthms have a better performance than the NAF algorthm over the whole SNR s and SNR r range Moreover, the proposed JSR-PG algorthm yelds the lowest MSE among all three algorthms The number of teratons requred for the JSR-PG algorthm to converge to ε = 0 3 n a typcal channel realzaton are lsted n Table, where we set K = 3 and SNR r = 0 db It can be seen that the JSR-PG algorthm converges wthn several teratons, and thus, t s realzable wth the advancement of modern chp desgn

6 Todng et al EURASIP Journal on Wreless Communcatons and Networkng 04, 04:5 Page 6 of NAF Algorthm ORO PG Algorthm JSR PG Algorthm Normalzed MSE SNRs db Fgure Example Normalzed MSE versus SNR s wth K = 3, SNR r = 0 db In the second example, we compare the bt error rate BER performance of the proposed JSR-PG algorthm n Algorthm, the ORO-PG algorthm n Algorthm, the suboptmal source and relay matrx desgn n [7], the one-way relay verson of the conjugate gradent-based source and relay algorthm n [6], and the NAF algorthm Fgure4dsplaysthesystemBERversusSNR s for a MIMO relay system wth K = 3 relay nodes and fxed SNR r at 0 db It can be seen from Fgure 4 that the proposed JSR-PG algorthm has a better BER performance than the exstng algorthms over the whole SNR s range In the thrd example, we study the effect of the number of relay nodes to the system BER performance usng the JSR-PG and ORO-PG algorthms Fgure 5 dsplays the system BER versus SNR s wth K =, 3, and 5 for fxed SNR r at 0 db It can be seen that at BER = 0,for both the ORO-PG algorthm and JSR-PG algorthm, we can acheve approxmately 3-dB gan by ncreasng from K = tok = 5 It can also be seen that the performance gan of the JSR-PG algorthm over the ORO-PG algorthm ncreases wth the ncreasng number of relay nodes 0 0 NAF Algorthm ORO PG Algorthm JSR PG Algorthm Normalzed MSE SNRr db Fgure 3 Example Normalzed MSE versus SNR r wth K = 3, SNR s = 0 db

7 Todng et al EURASIP Journal on Wreless Communcatons and Networkng 04, 04:5 Page 7 of 9 Table Iteratons requred untl convergence n the JSR-PG algorthm SNR s db Iteratons Conclusons In ths paper, we have derved the general structure of the optmal relay amplfyng matrces for lnear non-regeneratve MIMO relay communcaton systems wth multple relay nodes usng the projected gradent approach The proposed source and relay matrces mnmze the MSE of the sgnal waveform estmaton The smulaton results demonstrate that the proposed algorthm has mproved the MSE and BER performance compared wth exstng technques Appendces Appendx Proof of Theorem Wthout loss of generalty, F can be wrtten as F = [ V r, V r, ] [ ] [ ] A X U s, Y Z U, =,, K, s, 30 where V r, V r, = I Nr V r, V r,, s, U U s, = INr U s, U s,,suchthat V r, [ V r,, V r,] and Ūs, [ U s,, U ] s, are N r N r untary matrces Matrces A, X, Y, Z are arbtrary matrces wth dmensons of R r, R s,, R r, N r R s,, N r R r, R s,, N r R r, N r R s,,respectvely Substtutng 5 and 30 back nto 3, we obtan that rd, F sr, B = U r, r, A s, V s, and rd,f F rd, = U r, r, A A + X X r, U r, Thus, we can rewrte 3 as [ MSE tr I Nb + V s, s, A r, U r, = K U r, r, A A = + X X r, U r, + I N d ] U r, r, A s, V s, 3 = It can be seen that 3 s mnmzed by X = 0 Rr, N r R s,, =,, K Substtutng 5 and 30 back nto the left-hand sde of the transmsson power constrant 4, we have tr [ F sr, BB sr, + I ] N r F = tr A s, + I Rs, A + Y s, + I Rs, Y 3 +X X + Z Z, =,, K From 3, we fnd that X = 0 Rr, N r R s,, Y = 0 Nr R r, R s,,andz = 0 Nr R r, N r R s, mnmze the power consumpton at each relay node Thus, we have F = V r, A U s,, =,, K 0 0 BER NAF Algorthm ORO PG Algorthm Algorthm [7] Algorthm [6] JSR PG Algorthm Fgure 4 Example BER versus SNR s wth K = 3, SNR r = 0 db SNRs db

8 Todng et al EURASIP Journal on Wreless Communcatons and Networkng 04, 04:5 Page 8 of BER ORO PG Algorthm K= ORO PG Algorthm K=3 ORO PG Algorthm K=5 JSR PG Algorthm K= JSR PG Algorthm K=3 JSR PG Algorthm K=5 Fgure 5 Example 3 BER versus SNR s for dfferent K,SNR r = 0 db SNRs db Appendx Proof of Theorem Let us defne Z K j=,j = U r,j r,j A j s,j V s,j and Y Kj=,j = U r,j r,j A j A j r,j U r,j + I Nd Then,f A can be wrtten as [INb f A = tr + Z + V s, s, A r, U r, Y + U r, r, A A r, U r, Z + U r, r, A s, V ] s, 33 Applyng I Nb +A C A =INb A AA +C A, 33 can be wrtten as f A = tr I Nb Z +V s, s, A r, U r, [ Z + U r, r, A s, V s, Z +V s, s, A r, U r, 34 + Y + U r, r, A A r, U r, Z + U r, r, A s, V s, ] Let us now defne E Z + U r, r, A s, V s,, K Y + U r, r, A A r, U r,,andg E E + K Wecan rewrte 34 as f A = tr I Nb E G E = tr I Nb E E G 35 The dervatve of f A wth respect to A s gven by f A = tr E E G A A Z + U r, r, A s, V s, E = tr G E E G A + Y + U r, r, A A r, U r, tr E G U r, r, A s, V s, A 36 Defnng M G E E G, R U r, r,, S s, V s,,andd A r, U r,, we can rewrte 36 as f A A = tr M Z + U r, r, A s, V s, E A + M Y + U r, r, A D T E G s, U r, r, V T s, 37 = tr M R A S E + M R A D A T E G R S T 38 Fnally, the gradent of f A s gven by f A f A = A = M R T S E = R M E S + D T TS +M R T D T E G T R R G E S where stands for complex conjugate 39

9 Todng et al EURASIP Journal on Wreless Communcatons and Networkng 04, 04:5 Page 9 of 9 Competng nterests The authors declare that they have no competng nterests Acknowledgements The work of Yue Rong was supported n part by the Australan Research Councl s Dscovery Projects fundng scheme project number DP4003 The frst author Aprana Todng would lke to thank the gher Educaton Mnstry of Indonesa DIKTI and the Paulus Chrstan Unversty of Indonesa UKI-Paulus of Makassar, Indonesa, for provdng her wth a PhD scholarshp at Curtn Unversty, Perth, Australa Author detals Department of Electrcal Engneerng, Unverstas Krsten Indonesa Paulus, Jln Pernts Kemerdekaan No 8 Daya, Makassar 9043, Indonesa Department of Electronc and Electrcal Engneerng, Unversty College London, Gower Street, London WCE 7JE, UK 3 Department of Electrcal and Computer Engneerng, Curtn Unversty of Technology, Bentley, WA 60, Australa 0 S Boyd, L Vandenberghe, Convex Optmzaton Cambrdge Unversty Press, Cambrdge, 004 P Tseng, Further results on approxmatng nonconvex quadratc optmzaton by semdefnte programmng relaxaton SIAM J Optm 4, Y Nesterov, A Nemrovsk, Interor Pont Polynomal Algorthms n Convex Programmng, SIAM, Phladelpha, 994 do:086/ Cte ths artcle as: Todng et al: Jont source and relay desgn for MIMO mult-relay systems usng projected gradent approach EURASIP Journal on Wreless Communcatons and Networkng 04 04:5 Receved: 5 May 04 Accepted: September 04 Publshed: 5 September 04 References B Wang, J Zhang, A øst-madsen, On the capacty of MIMO relay channels IEEE Trans Inf Theory 5, X Tang, Y ua, Optmal desgn of non-regeneratve MIMO wreless relays IEEE Trans Wreless Commun 6, O Muñoz-Medna, J Vdal, A Agustín, Lnear transcever desgn n nonregeneratve relays wth channel state nformaton IEEE Trans Sgnal Process 55, W Guan, Luo, Jont MMSE transcever desgn n non-regeneratve MIMO relay systems IEEE Commun Lett, G L, Y Wang, T Wu, J uang, Jont lnear flter desgn n mult-user cooperatve non-regeneratve MIMO relay systems EURASIP J Wreless Commun Netw 009, Y Rong, Lnear non-regeneratve multcarrer MIMO relay communcatons based on MMSE crteron IEEE Trans Commun 58, Z Fang, Y ua, JC Koshy, Jont source and relay optmzaton for a nonregeneratve MIMO relay, n Proc IEEE Workshop Sensor Array Mult-Channel Sgnal Process Waltham, WA, USA, 4 July 006, pp Y Rong, X Tang, Y ua, A unfed framework for optmzng lnear non-regeneratve multcarrer MIMO relay communcaton systems IEEE Trans Sgnal Process 57, Y Rong, Y ua, Optmalty of dagonalzaton of mult-hop MIMO relays IEEE Trans Wreless Commun 8, Y Rong, Jont source and relay optmzaton for two-way lnear non-regeneratve MIMO relay communcatons IEEE Trans Sgnal Process 60, MRA Khandaker, Y Rong, Jont transcever optmzaton for multuser MIMO relay communcaton systems IEEE Trans Sgnal Process 60, MRA Khandaker, Y Rong, Jont source and relay optmzaton for multuser MIMO relay communcaton systems, n Proceedngs of the 4th Internatonal Conference on Sgnal Processng and Communcaton Systems Gold Coast, Australa, 3 5 Dec 00 3 Wan, W Chen, Jont source and relay desgn for multuser MIMO nonregeneratve relay networks wth drect lnks IEEE Trans Veh Technol 6, J Zeng, Z Chen, L L, Iteratve jont source and relay optmzaton for multuser MIMO relay systems, n Proceedngs of the IEEE Vehcular Technology Conference Quebec Cty, QC, Canada, 3 6 Sept 0 5 AS Behbahan, R Merched, AM Eltawl, Optmzatons of a MIMO relay network IEEE Trans Sgnal Process 56, C-C u, Y-F Chou, Precodng desgn of MIMO AF two-way multple-relay systems IEEE Sgnal Process Lett 0, A Todng, MRA Khandaker, Y Rong, Jont source and relay optmzaton for parallel MIMO relay networks EURASIP J Adv Sgnal Process 0, DP Bertsekas, Nonlnear Programmng, nd edn, Athena Scentfc, Belmont,999 9 KB Petersen, MS Petersen, The Matrx Cookbook dk/pubdb/pphp?374 Accessed 9 Sept 04 Submt your manuscrpt to a journal and beneft from: 7 Convenent onlne submsson 7 Rgorous peer revew 7 Immedate publcaton on acceptance 7 Open access: artcles freely avalable onlne 7 gh vsblty wthn the feld 7 Retanng the copyrght to your artcle Submt your next manuscrpt at 7 sprngeropencom

ECE559VV Project Report

ECE559VV Project Report ECE559VV Project Report (Supplementary Notes Loc Xuan Bu I. MAX SUM-RATE SCHEDULING: THE UPLINK CASE We have seen (n the presentaton that, for downlnk (broadcast channels, the strategy maxmzng the sum-rate

More information

Lecture Notes on Linear Regression

Lecture Notes on Linear Regression Lecture Notes on Lnear Regresson Feng L fl@sdueducn Shandong Unversty, Chna Lnear Regresson Problem In regresson problem, we am at predct a contnuous target value gven an nput feature vector We assume

More information

The Order Relation and Trace Inequalities for. Hermitian Operators

The Order Relation and Trace Inequalities for. Hermitian Operators Internatonal Mathematcal Forum, Vol 3, 08, no, 507-57 HIKARI Ltd, wwwm-hkarcom https://doorg/0988/mf088055 The Order Relaton and Trace Inequaltes for Hermtan Operators Y Huang School of Informaton Scence

More information

U.C. Berkeley CS294: Beyond Worst-Case Analysis Luca Trevisan September 5, 2017

U.C. Berkeley CS294: Beyond Worst-Case Analysis Luca Trevisan September 5, 2017 U.C. Berkeley CS94: Beyond Worst-Case Analyss Handout 4s Luca Trevsan September 5, 07 Summary of Lecture 4 In whch we ntroduce semdefnte programmng and apply t to Max Cut. Semdefnte Programmng Recall that

More information

Power Allocation/Beamforming for DF MIMO Two-Way Relaying: Relay and Network Optimization

Power Allocation/Beamforming for DF MIMO Two-Way Relaying: Relay and Network Optimization Power Allocaton/Beamformng for DF MIMO Two-Way Relayng: Relay and Network Optmzaton Je Gao, Janshu Zhang, Sergy A. Vorobyov, Ha Jang, and Martn Haardt Dept. of Electrcal & Computer Engneerng, Unversty

More information

MMA and GCMMA two methods for nonlinear optimization

MMA and GCMMA two methods for nonlinear optimization MMA and GCMMA two methods for nonlnear optmzaton Krster Svanberg Optmzaton and Systems Theory, KTH, Stockholm, Sweden. krlle@math.kth.se Ths note descrbes the algorthms used n the author s 2007 mplementatons

More information

Yong Joon Ryang. 1. Introduction Consider the multicommodity transportation problem with convex quadratic cost function. 1 2 (x x0 ) T Q(x x 0 )

Yong Joon Ryang. 1. Introduction Consider the multicommodity transportation problem with convex quadratic cost function. 1 2 (x x0 ) T Q(x x 0 ) Kangweon-Kyungk Math. Jour. 4 1996), No. 1, pp. 7 16 AN ITERATIVE ROW-ACTION METHOD FOR MULTICOMMODITY TRANSPORTATION PROBLEMS Yong Joon Ryang Abstract. The optmzaton problems wth quadratc constrants often

More information

Application of Nonbinary LDPC Codes for Communication over Fading Channels Using Higher Order Modulations

Application of Nonbinary LDPC Codes for Communication over Fading Channels Using Higher Order Modulations Applcaton of Nonbnary LDPC Codes for Communcaton over Fadng Channels Usng Hgher Order Modulatons Rong-Hu Peng and Rong-Rong Chen Department of Electrcal and Computer Engneerng Unversty of Utah Ths work

More information

Channel Estimation of MIMO Relay Systems With Multiple Relay Nodes

Channel Estimation of MIMO Relay Systems With Multiple Relay Nodes Receved September 21, 2017, accepted November 12, 2017, date of publcaton November 20, 2017, date of current verson December 22, 2017. Dgtal Object Identfer 10.1109/ACCESS.2017.2775202 Channel Estmaton

More information

Chapter 5. Solution of System of Linear Equations. Module No. 6. Solution of Inconsistent and Ill Conditioned Systems

Chapter 5. Solution of System of Linear Equations. Module No. 6. Solution of Inconsistent and Ill Conditioned Systems Numercal Analyss by Dr. Anta Pal Assstant Professor Department of Mathematcs Natonal Insttute of Technology Durgapur Durgapur-713209 emal: anta.bue@gmal.com 1 . Chapter 5 Soluton of System of Lnear Equatons

More information

Module 3 LOSSY IMAGE COMPRESSION SYSTEMS. Version 2 ECE IIT, Kharagpur

Module 3 LOSSY IMAGE COMPRESSION SYSTEMS. Version 2 ECE IIT, Kharagpur Module 3 LOSSY IMAGE COMPRESSION SYSTEMS Verson ECE IIT, Kharagpur Lesson 6 Theory of Quantzaton Verson ECE IIT, Kharagpur Instructonal Objectves At the end of ths lesson, the students should be able to:

More information

Lecture 20: Lift and Project, SDP Duality. Today we will study the Lift and Project method. Then we will prove the SDP duality theorem.

Lecture 20: Lift and Project, SDP Duality. Today we will study the Lift and Project method. Then we will prove the SDP duality theorem. prnceton u. sp 02 cos 598B: algorthms and complexty Lecture 20: Lft and Project, SDP Dualty Lecturer: Sanjeev Arora Scrbe:Yury Makarychev Today we wll study the Lft and Project method. Then we wll prove

More information

Robust transceiver design for AF MIMO relay systems with column correlations

Robust transceiver design for AF MIMO relay systems with column correlations Ttle Robust transcever desgn for AF MIMO relay systems wth column correlatons Author(s) Xng, C; Fe, Z; Wu, YC; Ma, S; Kuang, J Ctaton The 0 IEEE Internatonal Conference on Sgnal rocessng, Communcatons

More information

IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY, VOL. 66, NO. 10, OCTOBER

IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY, VOL. 66, NO. 10, OCTOBER IEEE TRANSACTIONS ON VEICULAR TECNOLOGY, VOL. 66, NO. 10, OCTOBER 2017 8979 Robust MMSE Transcever Desgn for Nonregeneratve Multcastng MIMO Relay Systems Lenn Gopal, Member, IEEE, Yue Rong, Senor Member,

More information

Solutions to exam in SF1811 Optimization, Jan 14, 2015

Solutions to exam in SF1811 Optimization, Jan 14, 2015 Solutons to exam n SF8 Optmzaton, Jan 4, 25 3 3 O------O -4 \ / \ / The network: \/ where all lnks go from left to rght. /\ / \ / \ 6 O------O -5 2 4.(a) Let x = ( x 3, x 4, x 23, x 24 ) T, where the varable

More information

On the spectral norm of r-circulant matrices with the Pell and Pell-Lucas numbers

On the spectral norm of r-circulant matrices with the Pell and Pell-Lucas numbers Türkmen and Gökbaş Journal of Inequaltes and Applcatons (06) 06:65 DOI 086/s3660-06-0997-0 R E S E A R C H Open Access On the spectral norm of r-crculant matrces wth the Pell and Pell-Lucas numbers Ramazan

More information

Linear Approximation with Regularization and Moving Least Squares

Linear Approximation with Regularization and Moving Least Squares Lnear Approxmaton wth Regularzaton and Movng Least Squares Igor Grešovn May 007 Revson 4.6 (Revson : March 004). 5 4 3 0.5 3 3.5 4 Contents: Lnear Fttng...4. Weghted Least Squares n Functon Approxmaton...

More information

LOW BIAS INTEGRATED PATH ESTIMATORS. James M. Calvin

LOW BIAS INTEGRATED PATH ESTIMATORS. James M. Calvin Proceedngs of the 007 Wnter Smulaton Conference S G Henderson, B Bller, M-H Hseh, J Shortle, J D Tew, and R R Barton, eds LOW BIAS INTEGRATED PATH ESTIMATORS James M Calvn Department of Computer Scence

More information

The Minimum Universal Cost Flow in an Infeasible Flow Network

The Minimum Universal Cost Flow in an Infeasible Flow Network Journal of Scences, Islamc Republc of Iran 17(2): 175-180 (2006) Unversty of Tehran, ISSN 1016-1104 http://jscencesutacr The Mnmum Unversal Cost Flow n an Infeasble Flow Network H Saleh Fathabad * M Bagheran

More information

Comparison of the Population Variance Estimators. of 2-Parameter Exponential Distribution Based on. Multiple Criteria Decision Making Method

Comparison of the Population Variance Estimators. of 2-Parameter Exponential Distribution Based on. Multiple Criteria Decision Making Method Appled Mathematcal Scences, Vol. 7, 0, no. 47, 07-0 HIARI Ltd, www.m-hkar.com Comparson of the Populaton Varance Estmators of -Parameter Exponental Dstrbuton Based on Multple Crtera Decson Makng Method

More information

Kernel Methods and SVMs Extension

Kernel Methods and SVMs Extension Kernel Methods and SVMs Extenson The purpose of ths document s to revew materal covered n Machne Learnng 1 Supervsed Learnng regardng support vector machnes (SVMs). Ths document also provdes a general

More information

x = x 1 + :::+ x K and the nput covarance matrces are of the form ± = E[x x y ]. 3.2 Dualty Next, we ntroduce the concept of dualty wth the followng t

x = x 1 + :::+ x K and the nput covarance matrces are of the form ± = E[x x y ]. 3.2 Dualty Next, we ntroduce the concept of dualty wth the followng t Sum Power Iteratve Water-fllng for Mult-Antenna Gaussan Broadcast Channels N. Jndal, S. Jafar, S. Vshwanath and A. Goldsmth Dept. of Electrcal Engg. Stanford Unversty, CA, 94305 emal: njndal,syed,srram,andrea@wsl.stanford.edu

More information

Hongyi Miao, College of Science, Nanjing Forestry University, Nanjing ,China. (Received 20 June 2013, accepted 11 March 2014) I)ϕ (k)

Hongyi Miao, College of Science, Nanjing Forestry University, Nanjing ,China. (Received 20 June 2013, accepted 11 March 2014) I)ϕ (k) ISSN 1749-3889 (prnt), 1749-3897 (onlne) Internatonal Journal of Nonlnear Scence Vol.17(2014) No.2,pp.188-192 Modfed Block Jacob-Davdson Method for Solvng Large Sparse Egenproblems Hongy Mao, College of

More information

Some Comments on Accelerating Convergence of Iterative Sequences Using Direct Inversion of the Iterative Subspace (DIIS)

Some Comments on Accelerating Convergence of Iterative Sequences Using Direct Inversion of the Iterative Subspace (DIIS) Some Comments on Acceleratng Convergence of Iteratve Sequences Usng Drect Inverson of the Iteratve Subspace (DIIS) C. Davd Sherrll School of Chemstry and Bochemstry Georga Insttute of Technology May 1998

More information

CIS526: Machine Learning Lecture 3 (Sept 16, 2003) Linear Regression. Preparation help: Xiaoying Huang. x 1 θ 1 output... θ M x M

CIS526: Machine Learning Lecture 3 (Sept 16, 2003) Linear Regression. Preparation help: Xiaoying Huang. x 1 θ 1 output... θ M x M CIS56: achne Learnng Lecture 3 (Sept 6, 003) Preparaton help: Xaoyng Huang Lnear Regresson Lnear regresson can be represented by a functonal form: f(; θ) = θ 0 0 +θ + + θ = θ = 0 ote: 0 s a dummy attrbute

More information

On the Multicriteria Integer Network Flow Problem

On the Multicriteria Integer Network Flow Problem BULGARIAN ACADEMY OF SCIENCES CYBERNETICS AND INFORMATION TECHNOLOGIES Volume 5, No 2 Sofa 2005 On the Multcrtera Integer Network Flow Problem Vassl Vasslev, Marana Nkolova, Maryana Vassleva Insttute of

More information

Some modelling aspects for the Matlab implementation of MMA

Some modelling aspects for the Matlab implementation of MMA Some modellng aspects for the Matlab mplementaton of MMA Krster Svanberg krlle@math.kth.se Optmzaton and Systems Theory Department of Mathematcs KTH, SE 10044 Stockholm September 2004 1. Consdered optmzaton

More information

Generalized Linear Methods

Generalized Linear Methods Generalzed Lnear Methods 1 Introducton In the Ensemble Methods the general dea s that usng a combnaton of several weak learner one could make a better learner. More formally, assume that we have a set

More information

Problem Set 9 Solutions

Problem Set 9 Solutions Desgn and Analyss of Algorthms May 4, 2015 Massachusetts Insttute of Technology 6.046J/18.410J Profs. Erk Demane, Srn Devadas, and Nancy Lynch Problem Set 9 Solutons Problem Set 9 Solutons Ths problem

More information

Vector Norms. Chapter 7 Iterative Techniques in Matrix Algebra. Cauchy-Bunyakovsky-Schwarz Inequality for Sums. Distances. Convergence.

Vector Norms. Chapter 7 Iterative Techniques in Matrix Algebra. Cauchy-Bunyakovsky-Schwarz Inequality for Sums. Distances. Convergence. Vector Norms Chapter 7 Iteratve Technques n Matrx Algebra Per-Olof Persson persson@berkeley.edu Department of Mathematcs Unversty of Calforna, Berkeley Math 128B Numercal Analyss Defnton A vector norm

More information

Estimation: Part 2. Chapter GREG estimation

Estimation: Part 2. Chapter GREG estimation Chapter 9 Estmaton: Part 2 9. GREG estmaton In Chapter 8, we have seen that the regresson estmator s an effcent estmator when there s a lnear relatonshp between y and x. In ths chapter, we generalzed the

More information

OPTIMAL COMBINATION OF FOURTH ORDER STATISTICS FOR NON-CIRCULAR SOURCE SEPARATION. Christophe De Luigi and Eric Moreau

OPTIMAL COMBINATION OF FOURTH ORDER STATISTICS FOR NON-CIRCULAR SOURCE SEPARATION. Christophe De Luigi and Eric Moreau OPTIMAL COMBINATION OF FOURTH ORDER STATISTICS FOR NON-CIRCULAR SOURCE SEPARATION Chrstophe De Lug and Erc Moreau Unversty of Toulon LSEET UMR CNRS 607 av. G. Pompdou BP56 F-8362 La Valette du Var Cedex

More information

Singular Value Decomposition: Theory and Applications

Singular Value Decomposition: Theory and Applications Sngular Value Decomposton: Theory and Applcatons Danel Khashab Sprng 2015 Last Update: March 2, 2015 1 Introducton A = UDV where columns of U and V are orthonormal and matrx D s dagonal wth postve real

More information

6.854J / J Advanced Algorithms Fall 2008

6.854J / J Advanced Algorithms Fall 2008 MIT OpenCourseWare http://ocw.mt.edu 6.854J / 18.415J Advanced Algorthms Fall 2008 For nformaton about ctng these materals or our Terms of Use, vst: http://ocw.mt.edu/terms. 18.415/6.854 Advanced Algorthms

More information

CRITICAL POINT ANALYSIS OF JOINT DIAGONALIZATION CRITERIA. Gen Hori and Jonathan H. Manton

CRITICAL POINT ANALYSIS OF JOINT DIAGONALIZATION CRITERIA. Gen Hori and Jonathan H. Manton CRITICAL POINT ANALYSIS OF JOINT DIAGONALIZATION CRITERIA Gen Hor and Jonathan H Manton Bran Scence Insttute, RIKEN, Satama 351-0198, Japan hor@bspbranrkengojp Department of Electrcal and Electronc Engneerng,

More information

Chapter - 2. Distribution System Power Flow Analysis

Chapter - 2. Distribution System Power Flow Analysis Chapter - 2 Dstrbuton System Power Flow Analyss CHAPTER - 2 Radal Dstrbuton System Load Flow 2.1 Introducton Load flow s an mportant tool [66] for analyzng electrcal power system network performance. Load

More information

Time-Varying Systems and Computations Lecture 6

Time-Varying Systems and Computations Lecture 6 Tme-Varyng Systems and Computatons Lecture 6 Klaus Depold 14. Januar 2014 The Kalman Flter The Kalman estmaton flter attempts to estmate the actual state of an unknown dscrete dynamcal system, gven nosy

More information

EURASIP Journal on Wireless Communications and Networking

EURASIP Journal on Wireless Communications and Networking EURASIP Journal on Wreless Communcatons and Networkng Ths Provsonal PDF corresponds to the artcle as t appeared upon acceptance. Fully formatted PDF and full text (TM) versons wll be made avalable soon.

More information

Research Article Green s Theorem for Sign Data

Research Article Green s Theorem for Sign Data Internatonal Scholarly Research Network ISRN Appled Mathematcs Volume 2012, Artcle ID 539359, 10 pages do:10.5402/2012/539359 Research Artcle Green s Theorem for Sgn Data Lous M. Houston The Unversty of

More information

Asymptotics of the Solution of a Boundary Value. Problem for One-Characteristic Differential. Equation Degenerating into a Parabolic Equation

Asymptotics of the Solution of a Boundary Value. Problem for One-Characteristic Differential. Equation Degenerating into a Parabolic Equation Nonl. Analyss and Dfferental Equatons, ol., 4, no., 5 - HIKARI Ltd, www.m-har.com http://dx.do.org/.988/nade.4.456 Asymptotcs of the Soluton of a Boundary alue Problem for One-Characterstc Dfferental Equaton

More information

IN cooperative communication systems, relay nodes can

IN cooperative communication systems, relay nodes can IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS, VOL 4, NO 0, OCTOBER 05 553 Tomlnson-Harashma Precodng Based Transcever Desgn for MIMO Relay Systems Wth Channel Covarance Informaton Lenn opal, Member, IEEE,

More information

University of Alberta. Library Release Form. Title of Thesis: Joint Bandwidth and Power Allocation in Wireless Communication Networks

University of Alberta. Library Release Form. Title of Thesis: Joint Bandwidth and Power Allocation in Wireless Communication Networks Unversty of Alberta Lbrary Release Form Name of Author: Xaowen Gong Ttle of Thess: Jont Bandwdth and Power Allocaton n Wreless Communcaton Networks Degree: Master of Scence Year ths Degree Granted: 2010

More information

Lecture 10 Support Vector Machines II

Lecture 10 Support Vector Machines II Lecture 10 Support Vector Machnes II 22 February 2016 Taylor B. Arnold Yale Statstcs STAT 365/665 1/28 Notes: Problem 3 s posted and due ths upcomng Frday There was an early bug n the fake-test data; fxed

More information

APPROXIMATE PRICES OF BASKET AND ASIAN OPTIONS DUPONT OLIVIER. Premia 14

APPROXIMATE PRICES OF BASKET AND ASIAN OPTIONS DUPONT OLIVIER. Premia 14 APPROXIMAE PRICES OF BASKE AND ASIAN OPIONS DUPON OLIVIER Prema 14 Contents Introducton 1 1. Framewor 1 1.1. Baset optons 1.. Asan optons. Computng the prce 3. Lower bound 3.1. Closed formula for the prce

More information

Errors for Linear Systems

Errors for Linear Systems Errors for Lnear Systems When we solve a lnear system Ax b we often do not know A and b exactly, but have only approxmatons  and ˆb avalable. Then the best thng we can do s to solve ˆx ˆb exactly whch

More information

Inexact Newton Methods for Inverse Eigenvalue Problems

Inexact Newton Methods for Inverse Eigenvalue Problems Inexact Newton Methods for Inverse Egenvalue Problems Zheng-jan Ba Abstract In ths paper, we survey some of the latest development n usng nexact Newton-lke methods for solvng nverse egenvalue problems.

More information

2E Pattern Recognition Solutions to Introduction to Pattern Recognition, Chapter 2: Bayesian pattern classification

2E Pattern Recognition Solutions to Introduction to Pattern Recognition, Chapter 2: Bayesian pattern classification E395 - Pattern Recognton Solutons to Introducton to Pattern Recognton, Chapter : Bayesan pattern classfcaton Preface Ths document s a soluton manual for selected exercses from Introducton to Pattern Recognton

More information

College of Computer & Information Science Fall 2009 Northeastern University 20 October 2009

College of Computer & Information Science Fall 2009 Northeastern University 20 October 2009 College of Computer & Informaton Scence Fall 2009 Northeastern Unversty 20 October 2009 CS7880: Algorthmc Power Tools Scrbe: Jan Wen and Laura Poplawsk Lecture Outlne: Prmal-dual schema Network Desgn:

More information

ISSN: ISO 9001:2008 Certified International Journal of Engineering and Innovative Technology (IJEIT) Volume 3, Issue 1, July 2013

ISSN: ISO 9001:2008 Certified International Journal of Engineering and Innovative Technology (IJEIT) Volume 3, Issue 1, July 2013 ISSN: 2277-375 Constructon of Trend Free Run Orders for Orthogonal rrays Usng Codes bstract: Sometmes when the expermental runs are carred out n a tme order sequence, the response can depend on the run

More information

4DVAR, according to the name, is a four-dimensional variational method.

4DVAR, according to the name, is a four-dimensional variational method. 4D-Varatonal Data Assmlaton (4D-Var) 4DVAR, accordng to the name, s a four-dmensonal varatonal method. 4D-Var s actually a drect generalzaton of 3D-Var to handle observatons that are dstrbuted n tme. The

More information

For now, let us focus on a specific model of neurons. These are simplified from reality but can achieve remarkable results.

For now, let us focus on a specific model of neurons. These are simplified from reality but can achieve remarkable results. Neural Networks : Dervaton compled by Alvn Wan from Professor Jtendra Malk s lecture Ths type of computaton s called deep learnng and s the most popular method for many problems, such as computer vson

More information

The Prncpal Component Transform The Prncpal Component Transform s also called Karhunen-Loeve Transform (KLT, Hotellng Transform, oregenvector Transfor

The Prncpal Component Transform The Prncpal Component Transform s also called Karhunen-Loeve Transform (KLT, Hotellng Transform, oregenvector Transfor Prncpal Component Transform Multvarate Random Sgnals A real tme sgnal x(t can be consdered as a random process and ts samples x m (m =0; ;N, 1 a random vector: The mean vector of X s X =[x0; ;x N,1] T

More information

Which Separator? Spring 1

Which Separator? Spring 1 Whch Separator? 6.034 - Sprng 1 Whch Separator? Mamze the margn to closest ponts 6.034 - Sprng Whch Separator? Mamze the margn to closest ponts 6.034 - Sprng 3 Margn of a pont " # y (w $ + b) proportonal

More information

The Synchronous 8th-Order Differential Attack on 12 Rounds of the Block Cipher HyRAL

The Synchronous 8th-Order Differential Attack on 12 Rounds of the Block Cipher HyRAL The Synchronous 8th-Order Dfferental Attack on 12 Rounds of the Block Cpher HyRAL Yasutaka Igarash, Sej Fukushma, and Tomohro Hachno Kagoshma Unversty, Kagoshma, Japan Emal: {garash, fukushma, hachno}@eee.kagoshma-u.ac.jp

More information

n α j x j = 0 j=1 has a nontrivial solution. Here A is the n k matrix whose jth column is the vector for all t j=0

n α j x j = 0 j=1 has a nontrivial solution. Here A is the n k matrix whose jth column is the vector for all t j=0 MODULE 2 Topcs: Lnear ndependence, bass and dmenson We have seen that f n a set of vectors one vector s a lnear combnaton of the remanng vectors n the set then the span of the set s unchanged f that vector

More information

Parametric fractional imputation for missing data analysis. Jae Kwang Kim Survey Working Group Seminar March 29, 2010

Parametric fractional imputation for missing data analysis. Jae Kwang Kim Survey Working Group Seminar March 29, 2010 Parametrc fractonal mputaton for mssng data analyss Jae Kwang Km Survey Workng Group Semnar March 29, 2010 1 Outlne Introducton Proposed method Fractonal mputaton Approxmaton Varance estmaton Multple mputaton

More information

Chapter Newton s Method

Chapter Newton s Method Chapter 9. Newton s Method After readng ths chapter, you should be able to:. Understand how Newton s method s dfferent from the Golden Secton Search method. Understand how Newton s method works 3. Solve

More information

Linear dispersion code with an orthogonal row structure for simplifying sphere decoding

Linear dispersion code with an orthogonal row structure for simplifying sphere decoding tle Lnear dsperson code wth an orthogonal row structure for smplfyng sphere decodng Author(s) Da XG; Cheung SW; Yuk I Ctaton he 0th IEEE Internatonal Symposum On Personal Indoor and Moble Rado Communcatons

More information

FUll-duplex (FD) is a promising technique to increase

FUll-duplex (FD) is a promising technique to increase 1 Jont Relay-User Beamformng Desgn n Full-Duplex Two-Way Relay Channel Zhgang Wen, Shua Wang, Xaoqng Lu, and Junwe Zou arxv:1610.04974v1 cs.it] 17 Oct 2016 Abstract A full-duplex two-way relay channel

More information

CSci 6974 and ECSE 6966 Math. Tech. for Vision, Graphics and Robotics Lecture 21, April 17, 2006 Estimating A Plane Homography

CSci 6974 and ECSE 6966 Math. Tech. for Vision, Graphics and Robotics Lecture 21, April 17, 2006 Estimating A Plane Homography CSc 6974 and ECSE 6966 Math. Tech. for Vson, Graphcs and Robotcs Lecture 21, Aprl 17, 2006 Estmatng A Plane Homography Overvew We contnue wth a dscusson of the major ssues, usng estmaton of plane projectve

More information

Multigradient for Neural Networks for Equalizers 1

Multigradient for Neural Networks for Equalizers 1 Multgradent for Neural Netorks for Equalzers 1 Chulhee ee, Jnook Go and Heeyoung Km Department of Electrcal and Electronc Engneerng Yonse Unversty 134 Shnchon-Dong, Seodaemun-Ku, Seoul 1-749, Korea ABSTRACT

More information

Solutions HW #2. minimize. Ax = b. Give the dual problem, and make the implicit equality constraints explicit. Solution.

Solutions HW #2. minimize. Ax = b. Give the dual problem, and make the implicit equality constraints explicit. Solution. Solutons HW #2 Dual of general LP. Fnd the dual functon of the LP mnmze subject to c T x Gx h Ax = b. Gve the dual problem, and make the mplct equalty constrants explct. Soluton. 1. The Lagrangan s L(x,

More information

Interactive Bi-Level Multi-Objective Integer. Non-linear Programming Problem

Interactive Bi-Level Multi-Objective Integer. Non-linear Programming Problem Appled Mathematcal Scences Vol 5 0 no 65 3 33 Interactve B-Level Mult-Objectve Integer Non-lnear Programmng Problem O E Emam Department of Informaton Systems aculty of Computer Scence and nformaton Helwan

More information

COMPARISON OF SOME RELIABILITY CHARACTERISTICS BETWEEN REDUNDANT SYSTEMS REQUIRING SUPPORTING UNITS FOR THEIR OPERATIONS

COMPARISON OF SOME RELIABILITY CHARACTERISTICS BETWEEN REDUNDANT SYSTEMS REQUIRING SUPPORTING UNITS FOR THEIR OPERATIONS Avalable onlne at http://sck.org J. Math. Comput. Sc. 3 (3), No., 6-3 ISSN: 97-537 COMPARISON OF SOME RELIABILITY CHARACTERISTICS BETWEEN REDUNDANT SYSTEMS REQUIRING SUPPORTING UNITS FOR THEIR OPERATIONS

More information

Design and Optimization of Fuzzy Controller for Inverse Pendulum System Using Genetic Algorithm

Design and Optimization of Fuzzy Controller for Inverse Pendulum System Using Genetic Algorithm Desgn and Optmzaton of Fuzzy Controller for Inverse Pendulum System Usng Genetc Algorthm H. Mehraban A. Ashoor Unversty of Tehran Unversty of Tehran h.mehraban@ece.ut.ac.r a.ashoor@ece.ut.ac.r Abstract:

More information

Matrix Approximation via Sampling, Subspace Embedding. 1 Solving Linear Systems Using SVD

Matrix Approximation via Sampling, Subspace Embedding. 1 Solving Linear Systems Using SVD Matrx Approxmaton va Samplng, Subspace Embeddng Lecturer: Anup Rao Scrbe: Rashth Sharma, Peng Zhang 0/01/016 1 Solvng Lnear Systems Usng SVD Two applcatons of SVD have been covered so far. Today we loo

More information

Norms, Condition Numbers, Eigenvalues and Eigenvectors

Norms, Condition Numbers, Eigenvalues and Eigenvectors Norms, Condton Numbers, Egenvalues and Egenvectors 1 Norms A norm s a measure of the sze of a matrx or a vector For vectors the common norms are: N a 2 = ( x 2 1/2 the Eucldean Norm (1a b 1 = =1 N x (1b

More information

The lower and upper bounds on Perron root of nonnegative irreducible matrices

The lower and upper bounds on Perron root of nonnegative irreducible matrices Journal of Computatonal Appled Mathematcs 217 (2008) 259 267 wwwelsevercom/locate/cam The lower upper bounds on Perron root of nonnegatve rreducble matrces Guang-Xn Huang a,, Feng Yn b,keguo a a College

More information

MIMO Systems and Channel Capacity

MIMO Systems and Channel Capacity MIMO Systems and Channel Capacty Consder a MIMO system wth m Tx and n Rx antennas. x y = Hx ξ Tx H Rx The power constrant: the total Tx power s x = P t. Component-wse representaton of the system model,

More information

On an Extension of Stochastic Approximation EM Algorithm for Incomplete Data Problems. Vahid Tadayon 1

On an Extension of Stochastic Approximation EM Algorithm for Incomplete Data Problems. Vahid Tadayon 1 On an Extenson of Stochastc Approxmaton EM Algorthm for Incomplete Data Problems Vahd Tadayon Abstract: The Stochastc Approxmaton EM (SAEM algorthm, a varant stochastc approxmaton of EM, s a versatle tool

More information

5 The Rational Canonical Form

5 The Rational Canonical Form 5 The Ratonal Canoncal Form Here p s a monc rreducble factor of the mnmum polynomal m T and s not necessarly of degree one Let F p denote the feld constructed earler n the course, consstng of all matrces

More information

Additional Codes using Finite Difference Method. 1 HJB Equation for Consumption-Saving Problem Without Uncertainty

Additional Codes using Finite Difference Method. 1 HJB Equation for Consumption-Saving Problem Without Uncertainty Addtonal Codes usng Fnte Dfference Method Benamn Moll 1 HJB Equaton for Consumpton-Savng Problem Wthout Uncertanty Before consderng the case wth stochastc ncome n http://www.prnceton.edu/~moll/ HACTproect/HACT_Numercal_Appendx.pdf,

More information

8.4 COMPLEX VECTOR SPACES AND INNER PRODUCTS

8.4 COMPLEX VECTOR SPACES AND INNER PRODUCTS SECTION 8.4 COMPLEX VECTOR SPACES AND INNER PRODUCTS 493 8.4 COMPLEX VECTOR SPACES AND INNER PRODUCTS All the vector spaces you have studed thus far n the text are real vector spaces because the scalars

More information

Comparison of Wiener Filter solution by SVD with decompositions QR and QLP

Comparison of Wiener Filter solution by SVD with decompositions QR and QLP Proceedngs of the 6th WSEAS Int Conf on Artfcal Intellgence, Knowledge Engneerng and Data Bases, Corfu Island, Greece, February 6-9, 007 7 Comparson of Wener Flter soluton by SVD wth decompostons QR and

More information

A Fast Computer Aided Design Method for Filters

A Fast Computer Aided Design Method for Filters 2017 Asa-Pacfc Engneerng and Technology Conference (APETC 2017) ISBN: 978-1-60595-443-1 A Fast Computer Aded Desgn Method for Flters Gang L ABSTRACT *Ths paper presents a fast computer aded desgn method

More information

Speeding up Computation of Scalar Multiplication in Elliptic Curve Cryptosystem

Speeding up Computation of Scalar Multiplication in Elliptic Curve Cryptosystem H.K. Pathak et. al. / (IJCSE) Internatonal Journal on Computer Scence and Engneerng Speedng up Computaton of Scalar Multplcaton n Ellptc Curve Cryptosystem H. K. Pathak Manju Sangh S.o.S n Computer scence

More information

Statistical inference for generalized Pareto distribution based on progressive Type-II censored data with random removals

Statistical inference for generalized Pareto distribution based on progressive Type-II censored data with random removals Internatonal Journal of Scentfc World, 2 1) 2014) 1-9 c Scence Publshng Corporaton www.scencepubco.com/ndex.php/ijsw do: 10.14419/jsw.v21.1780 Research Paper Statstcal nference for generalzed Pareto dstrbuton

More information

MACHINE APPLIED MACHINE LEARNING LEARNING. Gaussian Mixture Regression

MACHINE APPLIED MACHINE LEARNING LEARNING. Gaussian Mixture Regression 11 MACHINE APPLIED MACHINE LEARNING LEARNING MACHINE LEARNING Gaussan Mture Regresson 22 MACHINE APPLIED MACHINE LEARNING LEARNING Bref summary of last week s lecture 33 MACHINE APPLIED MACHINE LEARNING

More information

SIO 224. m(r) =(ρ(r),k s (r),µ(r))

SIO 224. m(r) =(ρ(r),k s (r),µ(r)) SIO 224 1. A bref look at resoluton analyss Here s some background for the Masters and Gubbns resoluton paper. Global Earth models are usually found teratvely by assumng a startng model and fndng small

More information

The Geometry of Logit and Probit

The Geometry of Logit and Probit The Geometry of Logt and Probt Ths short note s meant as a supplement to Chapters and 3 of Spatal Models of Parlamentary Votng and the notaton and reference to fgures n the text below s to those two chapters.

More information

The Study of Teaching-learning-based Optimization Algorithm

The Study of Teaching-learning-based Optimization Algorithm Advanced Scence and Technology Letters Vol. (AST 06), pp.05- http://dx.do.org/0.57/astl.06. The Study of Teachng-learnng-based Optmzaton Algorthm u Sun, Yan fu, Lele Kong, Haolang Q,, Helongang Insttute

More information

Estimating the Fundamental Matrix by Transforming Image Points in Projective Space 1

Estimating the Fundamental Matrix by Transforming Image Points in Projective Space 1 Estmatng the Fundamental Matrx by Transformng Image Ponts n Projectve Space 1 Zhengyou Zhang and Charles Loop Mcrosoft Research, One Mcrosoft Way, Redmond, WA 98052, USA E-mal: fzhang,cloopg@mcrosoft.com

More information

Natural Language Processing and Information Retrieval

Natural Language Processing and Information Retrieval Natural Language Processng and Informaton Retreval Support Vector Machnes Alessandro Moschtt Department of nformaton and communcaton technology Unversty of Trento Emal: moschtt@ds.untn.t Summary Support

More information

The Concept of Beamforming

The Concept of Beamforming ELG513 Smart Antennas S.Loyka he Concept of Beamformng Generc representaton of the array output sgnal, 1 where w y N 1 * = 1 = w x = w x (4.1) complex weghts, control the array pattern; y and x - narrowband

More information

Sum Capacity of Multiuser MIMO Broadcast Channels with Block Diagonalization

Sum Capacity of Multiuser MIMO Broadcast Channels with Block Diagonalization Sum Capacty of Multuser MIMO Broadcast Channels wth Block Dagonalzaton Zukang Shen, Runhua Chen, Jeffrey G. Andrews, Robert W. Heath, Jr., and Bran L. Evans The Unversty of Texas at Austn, Austn, Texas

More information

VQ widely used in coding speech, image, and video

VQ widely used in coding speech, image, and video at Scalar quantzers are specal cases of vector quantzers (VQ): they are constraned to look at one sample at a tme (memoryless) VQ does not have such constrant better RD perfomance expected Source codng

More information

Tracking with Kalman Filter

Tracking with Kalman Filter Trackng wth Kalman Flter Scott T. Acton Vrgna Image and Vdeo Analyss (VIVA), Charles L. Brown Department of Electrcal and Computer Engneerng Department of Bomedcal Engneerng Unversty of Vrgna, Charlottesvlle,

More information

Power Allocation for Distributed BLUE Estimation with Full and Limited Feedback of CSI

Power Allocation for Distributed BLUE Estimation with Full and Limited Feedback of CSI Power Allocaton for Dstrbuted BLUE Estmaton wth Full and Lmted Feedback of CSI Mohammad Fanae, Matthew C. Valent, and Natala A. Schmd Lane Department of Computer Scence and Electrcal Engneerng West Vrgna

More information

Resource Allocation with a Budget Constraint for Computing Independent Tasks in the Cloud

Resource Allocation with a Budget Constraint for Computing Independent Tasks in the Cloud Resource Allocaton wth a Budget Constrant for Computng Independent Tasks n the Cloud Wemng Sh and Bo Hong School of Electrcal and Computer Engneerng Georga Insttute of Technology, USA 2nd IEEE Internatonal

More information

Lossy Compression. Compromise accuracy of reconstruction for increased compression.

Lossy Compression. Compromise accuracy of reconstruction for increased compression. Lossy Compresson Compromse accuracy of reconstructon for ncreased compresson. The reconstructon s usually vsbly ndstngushable from the orgnal mage. Typcally, one can get up to 0:1 compresson wth almost

More information

Lecture 17: Lee-Sidford Barrier

Lecture 17: Lee-Sidford Barrier CSE 599: Interplay between Convex Optmzaton and Geometry Wnter 2018 Lecturer: Yn Tat Lee Lecture 17: Lee-Sdford Barrer Dsclamer: Please tell me any mstake you notced. In ths lecture, we talk about the

More information

Linear precoding via conic optimization for fixed MIMO receivers

Linear precoding via conic optimization for fixed MIMO receivers Lnear precodng va conc optmzaton for fxed MIMO recevers Am Wesel, Yonna C Eldar, and Shlomo Shama (Shtz) Department of Electrcal Engneerng Technon - Israel Insttute of Technology June 8, 004 Abstract We

More information

Temperature. Chapter Heat Engine

Temperature. Chapter Heat Engine Chapter 3 Temperature In prevous chapters of these notes we ntroduced the Prncple of Maxmum ntropy as a technque for estmatng probablty dstrbutons consstent wth constrants. In Chapter 9 we dscussed the

More information

Feature Selection: Part 1

Feature Selection: Part 1 CSE 546: Machne Learnng Lecture 5 Feature Selecton: Part 1 Instructor: Sham Kakade 1 Regresson n the hgh dmensonal settng How do we learn when the number of features d s greater than the sample sze n?

More information

STAT 309: MATHEMATICAL COMPUTATIONS I FALL 2018 LECTURE 16

STAT 309: MATHEMATICAL COMPUTATIONS I FALL 2018 LECTURE 16 STAT 39: MATHEMATICAL COMPUTATIONS I FALL 218 LECTURE 16 1 why teratve methods f we have a lnear system Ax = b where A s very, very large but s ether sparse or structured (eg, banded, Toepltz, banded plus

More information

Notes on Frequency Estimation in Data Streams

Notes on Frequency Estimation in Data Streams Notes on Frequency Estmaton n Data Streams In (one of) the data streamng model(s), the data s a sequence of arrvals a 1, a 2,..., a m of the form a j = (, v) where s the dentty of the tem and belongs to

More information

The Exact Formulation of the Inverse of the Tridiagonal Matrix for Solving the 1D Poisson Equation with the Finite Difference Method

The Exact Formulation of the Inverse of the Tridiagonal Matrix for Solving the 1D Poisson Equation with the Finite Difference Method Journal of Electromagnetc Analyss and Applcatons, 04, 6, 0-08 Publshed Onlne September 04 n ScRes. http://www.scrp.org/journal/jemaa http://dx.do.org/0.46/jemaa.04.6000 The Exact Formulaton of the Inverse

More information

ANSWERS. Problem 1. and the moment generating function (mgf) by. defined for any real t. Use this to show that E( U) var( U)

ANSWERS. Problem 1. and the moment generating function (mgf) by. defined for any real t. Use this to show that E( U) var( U) Econ 413 Exam 13 H ANSWERS Settet er nndelt 9 deloppgaver, A,B,C, som alle anbefales å telle lkt for å gøre det ltt lettere å stå. Svar er gtt . Unfortunately, there s a prntng error n the hnt of

More information

A Hybrid Variational Iteration Method for Blasius Equation

A Hybrid Variational Iteration Method for Blasius Equation Avalable at http://pvamu.edu/aam Appl. Appl. Math. ISSN: 1932-9466 Vol. 10, Issue 1 (June 2015), pp. 223-229 Applcatons and Appled Mathematcs: An Internatonal Journal (AAM) A Hybrd Varatonal Iteraton Method

More information

LINEAR REGRESSION ANALYSIS. MODULE IX Lecture Multicollinearity

LINEAR REGRESSION ANALYSIS. MODULE IX Lecture Multicollinearity LINEAR REGRESSION ANALYSIS MODULE IX Lecture - 31 Multcollnearty Dr. Shalabh Department of Mathematcs and Statstcs Indan Insttute of Technology Kanpur 6. Rdge regresson The OLSE s the best lnear unbased

More information