Dynamic Method of Network Coding Based Retransmission for Wireless Multicast
|
|
- Sara Thompson
- 6 years ago
- Views:
Transcription
1 Dynamc Method of Networ Codng Based Retransmsson for Wreless Multcast Houy L State Key Lab of ISN Xdan Unversty X an Chna Emal: lhouy2013@hotmal.com Yng L State Key Lab of ISN Xdan Unversty X an Chna yl@mal.xdan.edu.com Abstract In wreless sngle-hop multcast wth networ codng based Automatc Repeat request (NC-ARQ) scheme, the sender needs to combne one or several lost pacets as one retransmsson pacet to reduce the number of multcast transmssons. Ths paper proposes a new scheme, called Dynamc Networ Codng based ARQ (DNC-ARQ), to mprove the effcency of the multcast transmssons. The man dea s to allow the BS to select the combnaton for the lost pacets n dynamc range: the BS would not retransmt for a lost pacet untl an optmum combnaton s found or the lost pacet acheve the mum delay. We also derve the upper and low bound of DNC-ARQ scheme and gve a practcal algorthm. Both theoretcal analyss and smulaton results confrm the proposed scheme hold more effcent and less decodng delay than the tradtonal NC-ARQ. I. INTRODUCTION In a multcast transmsson system, a sender needs to dssemnate dentcal nformaton to a group of recevers. To ensure relablty, two man approaches are adopted: Forward Error Correcton (FEC) and Automatc Repeat request (ARQ). In multmeda broadcastng and multcast servce (MBMS) system [1], two nds of FEC, Turbo code and Raptor code, wor the physcal layer and applcaton layer respectvely. Both of them need to broadcast the orgnal nformaton and the correspondng redundant nformaton to the recevers. If the amount of the lost data s suffcently small, the recever can recover the lost data relably va usng proper decodng algorthm. When ARQ scheme s employed, the sender wll rebroadcast the lost pacets. However, the transmsson effcency of such system would be very low when dfferent users lost dfferent pacets. In such case, the sender has to broadcast the lost pacet one by one. To ncrease the transmsson effcency, Salm Y. et al. [2] proposed a method to construct the retransmsson pacet by usng the lnear combnaton of several lost pacets. They also proved that the problem of fndng the optmal combnaton of the lost pacets to mnmze the number of retransmsson pacets s a NP-complete problem. S. Katt et al. [3] showed that the lnear combnaton approach for sngle-hop relable multcast can be consdered as a nd of networ codng, named as Opportunstc Networ Codng (ONC). Recently, the combnaton of networ codng wth ARQ scheme has nspred consderable nterests snce ths scheme can utlze the transmsson effcently. In ths paper, we call ths scheme as networ codng based ARQ (NC-ARQ). Nguyen et al. [4] proposed a class of NC-ARQ schemes. In these schemes, the sender wats untl a buffer of pacets have been transmtted before any retransmsson taes place. Durng the retransmsson phase, the sender forms a new pacet by combnng some of the lost pacets. The transmsson effcency of the proposed scheme s also analyzed n [4]. Snce the number of the pacets durng the transmsson phase s fxed, we call ths scheme as fxed-sze networ codng based ARQ (FNC-ARQ). The algorthm on the combnaton of the lost pacets are dscussed n [5] and [6] respectvely. Ch et al. also gave two practcable FNC-ARQ schemes [7]. Both transmsson effcency and transmsson delay are analyzed va computer smulaton.the comparson between the FEC scheme wth fountan code and the FNC-ARQ scheme s conducted by Nguyen et al. [8]. It s demonstrated that FNC- ARQ performs better than FEC scheme wth fountan code n the case of small scale of users and vce versa n the case of large scale. Meanwhle, some Networ Codng based HARQ (NC-HARQ) schemes have also been explored [9][10]. In the FNC-ARQ scheme, all orgnal pacets are transmtted by buffer and buffer. So the sender can t select combnaton for the lost pacet between dfferent buffers. That mght loss some transmsson effcency. We manage to desgn a new scheme to mze the effcency of every retransmsson pacet and eep the decodng delay at the same level wth FNC-ARQ scheme. So we buld the math model of NC-ARQ wth the scenaro of Wreless Sngle-Hop Relable Multcast (WSHRM). Then we summarze some characterstcs of NC- ARQ and propose a new class of NC-ARQ: Dynamc Networ Codng based ARQ (DNC-ARQ) whch replaces the fxed-sze buffer wth a dynamc lst. We also desgn a codng selecton algorthm for DNC-ARQ and show that DNC-ARQ scheme has better transmsson effcency than FNC-ARQ. II. MATH MODEL A wreless sngle-hop multcast transmsson system contans two parts: one s the Base Staton (BS), the sender; and the other s a group of User Equpment (UE), the recevers. Let M be the number of UEs n the group. The UE group can be wrtten as U = {U 1,..., U M }.Let H denote the number of retransmsson pacets. And the L orgnal pacets can be wrtten as P = {P1 0,..., PL 0 }. In dfferent NC-ARQ scheme, these L + H pacets are transmtted n dfferent order. We
2 wrte j th pacet broadcasted by the BS as P j = c 1,j P 0 1 c 2,j P c L,j P 0 L where 1 j L + H. It can be easly seen that c,j can be ether 0 or 1. The pacet P j can be represented by φ j = [c 1,j, c 2,j,..., c L,j ] whch s called as the networ codng vector (NCV). If P j s an orgn pacet, only one element n the NCV s 1 and all other element are 0. Thus we have: c 0 j = L g=1 c g,j = 1.If P j s a retransmsson pacet, we have:c 0 j = L g=1 c g,j 1. The pacet P j receved by U can be also consdered as a vector φ j,. If U receve the pacet P j successfully, φ j, = φ j. Otherwse, φ j, = 0. We call φ j, as j th decodng vector of U. After the P j has been transmtted, the BS holds j NCVs. We can wrte them as a matrx: Φ j = [φ 1,..., φ j ] T We call ths matrx as networ codng matrx for j th slot. Lewse, the U holds j decodng vectors at j th slot. We can wrte them as a decodng matrx of U for current slot: Φ j, = [φ 1,,..., φ j, ] T When P j+1 has been transmtted, the networ codng matrx of BS s updated as: Φ j+1 = [Φ j, φ j+1 ] T, and decodng matrx for U s updated as: Φ j+1, = [Φ j,, φ j+1, ] T. III. DNC-ARQ SCHEME In ths chapter, we frstly analyze the characterstcs of the optmal retransmsson pacets. And then based on these characterstcs, we propose the DNC-ARQ scheme. An example s also gven to show the dfference between DNC-ARQ and FNC-ARQ. A. Characters Based on the WSHRM model, we frst defne a parameter to descrbe the effcency of one retransmsson pacet. Defnton 1. Sngle Pacet Effcency (SPE) Supposng the NCV of P j s φ j, the appendx set of φ j s: θ j = { Φ j 1, s lnearly ndependence of φ j } And the SPE of the pacet can be expressed as: µ j = θ j /M Obvously, we have θ j M and µ j 1, whch means the most effcency pacet s lnearly ndependent of these decodng matrxes of all UEs. The defnton also suggests that a pacet s conceved to be more effcent than the other f ts nformaton s meanngful to more UEs. The concept that a pacet s conceved to be meanngful to a UE means that ths pacet can t be expressed by a combnaton of those pacets receved successfully by ths UE. We gve the followng defnton to descrbe the most effcent pacet. Defnton 2. Perfect Match Pacet (PMP) We call P j as perfect match pacet, f t satsfes: θ j = M Fg. 1. An example of dynamc lst Defnton 3. Maxmum Match Pacet (MMP) Supposng that the BS has transmtted j 1 pacets ncludng t j orgnal pacets, we can wrte another auxlary set: ϑ j = { Ran(Φ j 1, ) < t j }.We call P j as mum match pacet, f t satsfes: θ j = ϑ j Apparently, the SPE of a PMP s 1. And that of a MMP s µ = ϑ j /M 1. The orgn pacet transmtted n frst tme can be consdered as a PMP. But retransmsson pacet may nether a PMP nor a MMP. B. Process of DNC-ARQ To mze the SPE of the retransmsson pacet, we propose a new class NC-ARQ, Dynamc Networ Codng based ARQ (DNC-ARQ) scheme. Instead of employng the fxed sze buffer, the DNC-ARQ adopt a dynamc lst whch s created and updated by the BS. The BS retransmts a pacet only f the pacet s a PMP or a MMP. We frstly ntroduce the concepton of dynamc lst and then show the process of DNC-ARQ scheme. The dynamc lst consst of several dynamc chans, each of whch contans the basc nformaton of certan orgn pacet. The dynamc chan for Pj 0 can be wrtten as followng: CH j = {NP j, ST j, E j }, {TN j, } where NP j s the ID of orgnal pacet Pj 0. In ths paper, we let NP j = j. ST j s the survve tme of Pj 0. And {TN j,} s the ID set of those UEs whch can t recover Pj 0. Supposng j orgnal pacets and h retrans-pacets have been transmtted, the set {TN j, } can be wrtten as {TN j, } = { Φ h+j, (:, j) = 0, 1 M} Let E j denote the element number of {TN j, }. We have E j = {TN j, }. When all UEs can recover the orgnal pacet Pj 0 ( E j = 0), the BS wll delete the CH j from the dynamc lst. Fg.2. llustrates an example for the dynamc lst. The decodng delay s also consdered n dynamc lst. We denote the mum affordable decodng delay of the system as NH unt tme. ST j s ntalzed as 1, whch means the
3 orgnal pacet Pj 0 has been just transmtted by BS. ST j wll be ncreased by 1, once he BS multcast orgn or retransmsson pacet. When ST j s approachng to NH, the BS wll multcast retransmsson pacets to mae all UEs recover Pj 0 as soon as possble. In ths way, we control the decodng delay of DNC-ARQ scheme n an acceptable level. Now, we descrbe the DNC-ARQ scheme step by step: Step1: Intalzng dynamc lst. The BS multcast ρ orgnal pacets one by one, and then ntal the dynamc lst accordng to the feedbac message from M UEs. The dynamc lst created by the BS can be wrtten as: DL = [CH 1 ;...; CH ρ ] = [{NP 1, ST 1, E 1 }{TN 1, };...; {NP ρ, ST ρ, E ρ }{TN ρ, }] where ρ = NH/2.And then BS delete the CH λ whose E λ = 0 (1 λ ρ). Step2: Multcastng pacet. There are two cases n ths step: multcastng an orgnal pacet or a retransmsson pacet. We suppose that DL n current slot nclude y dynamc chans, whch can be wrtten as [CH x ;...; CH x+y ]. If ST x < NH-1 and BS can t fnd a PMP for {TN j, }, the BS multcast the next orgnal pacet. Otherwse, the BS multcast a retransmsson pacet. If a PMP s found for {TN j, } whle ST x < NH 1, the BS wll multcast the PMP to UEs. If ST x NH 1, the BS wll see a MMP for {TN j, } and multcast t to UEs. Both the PMP and the MMP can be wrtten as: P r = φ r (1)P 0 1 φ r (2)P φ r (L)P 0 L where φ r s the NCV of the retransmsson pacet. Step3: Recevng feedbac. The BS receves ACK/NACK message from UEs, and create a success set γ s to mar the ID of UEs from whch BS receves ACK message. Supposng the BS receves n ACKs from UEs, the set can be wrtten as: γ s = [TN 1 ac,..., TN n ac]. Step4: Updatng dynamc lst. Correspondng to the step 2, there are two cases n ths step: updatng dynamc lst for orgnal pacet or for retrans-pacet. In both two case, the BS wll let all ST from CH x to CH x+y ncreased by 1 at frst. For orgnal pacet, the BS wll create a new dynamc chan CH cs = {NP cs, 1, E cs }{TN cs, } where cs [x+y+1,..., x+y-1+st x+y ], NP cs = cs, {TN cs, } = { / γ s, 1 M}. For retransmsson pacet, the BS deletes some elements n {TN x, },..., {TN x+y, }, accordng to the success set γ s and networ codng vector φ r. More specfcally, the TN j, s deleted both from dynamc lst and γ s, f φ r (j) = 1 and TN j, γ s. After updatng the dynamc lst, the BS removes those dynamc chan whose E = 0 and goes bac to step 2 untl L orgnal have been recovered by M UEs. C. Example Here we tae an example to llustrate the dfference of the two schemes. Fg.2. s called as the feedbac matrx of NC- ARQ scheme. Any lne of ths matrx note the ACK/NACK Fg. 2. Feedbac Matrx of NC-ARQ schemes message of one UE for all orgnal pacets, where 1 means ACK and 0 means NCAK. We have 10 orgnal pacets and 4 UEs n ths example. In FNC-ARQ scheme wth buffer sze 5, the BS frstly multcast the pacets from P 1 to P 5 one by one, and then multcasts the followng retransmsson pacets to ensure all UEs recover ther lost pacets: R 1 = P 1 P 2 P 3, R 2 = P 4, R 3 = P 5 Then the BS multcast the pacets from P 6 to P 10 one by one, and then multcast the retransmsson pacets: R 4 = P 6, R 5 = P 7, R 6 = P 10.Obvously, the number of the retransmsson pacets n a buffer s equal to the largest number of lost pacets for one UE n the buffer. And the average SPE of those 6 retransmsson pacets s In DNC-ARQ scheme where the survve tme s equal to 5, we could fnd two prefect match retransmsson pacets (the yellow and orange area) and two mum match retransmsson pacets (the blue and gray area).they are: R 1 = P 1 P 2 P 3, R 2 = P 4 P 6, R 3 = P 5 P 7, R 4 = P 10. And the average SPE of those 4 retransmsson pacets s In order to explan the dfference between the FNC-ARQ and DNC-ARQ, we dvde the matrx above nto two buffers, each of whch has sze 5. In buffer 1, UE 1, UE 2, UE 3 and UE 4 loss 1,1,1,3 pacets respectvely. Thus three retransmsson pacets are need to recover all lost pacets n buffer 1. They are the pacets: R 1, R 2 and R 3. In buffer 2, UE 1, UE 2, UE 3 and UE 4 loss 1,3,3,1 pacets respectvely. But R 2 recover P6 0 for UE 1, UE 2, UE 3. R 3 recover P7 0 for UE 2, UE 3. So the largest number of lost pacets s 1 and just one retransmsson pacets s needed to buffer 2. It s pacet R 4. As we have seen n ths example, the DNC-ARQ scheme also follow the law that the number of retransmsson pacets n a buffer s equal to the largest number of lost pacets for one UE n the buffer. But n DNC-ARQ scheme, the retransmsson pacet could be a combnaton of lost pacets from dfferent buffers. That means the retransmsson pacets of last buffer mght recover some lost pacets n ths buffer. So n DNC- ARQ scheme, we count the largest number of lost pacets after subtractng those recovered pacets. D. Maxmum match algorthm Prefect match for {TN j, } means we could fnd a networ codng vector whch satsfes: { 1 f s = x φ r (s) = (3.1) 0 f s < x or s > x + y
4 In fact, α s a dscrete random varable whose probablty densty functon depends on the pacet loss probablty p and L. We can denote as: α f α (x, p, L) Fg. 3. Maxmum/Perfect Macth Algorthm θ r = M, where θ r s appendx set of φ r (3.2) Maxmum match for {TN j, } means we could fnd a Networ codng vector whch satsfes (3.1) and θ r = ϑ r, (3.3) Supposng DL ncludes y dynamc chans: [CH x ;...; CH x+y ], the procedure of MMA s shown n Fg.3. IV. THEORETICAL ANALYSIS The transmsson effcency, decodng delay and decodng complexty are three most mportant factors n WSHRM system. A good NC-ARQ scheme s the one that can mze the transmsson effcency, uder the lmt of transmsson delay and decodng complexty. We theoretcally analyze both transmsson effcency and decodng delay of DNC-ARQ n ths secton. A. transmsson effcency Defnton 4. Transmsson Bandwdth(TB) s the rato between the number of pacets needed to be transmtted and the number of the orgnal pacets: TB = (L + H)/L (4.1) Obvously, the TB s the bandwdth that should be occuped to multcast L pacets to M UEs. TB s also adopted by many other research about NC-ARQ scheme. Lemma 1. Supposng α 1,..., α M s the number of lost pacets for UE 1,..., UE M respectvely, H (α 1,..., α M ). Proof: For an arbtrary user equpment UE, the decodng matrx for L+H slot s Φ L+H, = [φ 1,,..., φ L+H, ]. UE loss α pacets, whch means α decodng vectors of φ L+H, are equal to 0. Because UE can recover L orgnal pacets, ran(φ L+H, ) L. We have: L + H α L. Thus, H (α 1,..., α M ) (4.2) The equalty s hold, f the NCV n networ codng matrx Φ L+H satsfy: φ j (s) can be 1 or 0, for any j [1,..., L + H] and any s [1,..., L]. Accordng to [4], the probablty densty functon of α s: p x (1 p ) N x CN z Cz 1 x 1, wthx N z=0 f α (x, p, L) = p x (1 p ) N N C y N Cz 1 x 1, wthx > N z=0 (4.3) where x s the value of α, p s the pacet loss probablty of UE. In deal crcumstance, H = M =1 α. So the H can be also consder as a dscrete random varable. And we can ts probablty densty functon H = M α f H (y, p, M, L) =1 M y M y 1 = f α (x, p, L) f α (x, p, L) =1 x=0 =1 x=0 (4.4) where y s the value of H, p = [p 1,..., p M ], and M s the number of UE, L s the number of orgnal pacets. Theorem 1. The upper bound of TB wth DNC-ARQ s: where H f H(y, p, M, NH) TB DN C 1 + E(H ) / NH (4.5) Proof: To analyze convenently, we devce the L orgnal pacets n l buffers,each of whch holds NH orgnal pacets. We frstly consder the crcumstance wthout nteracton between dfferent buffers. For -th buffer, the number of lost pacets of UE s noted as α. Supposng that the BS multcast H retransmsson pacets for th buffer, we haveh (α ). In DNC-ARQ scheme, the BS can search combnaton for lost pacets n last NH-1 and next NH-1 pacets. So all NCVs n ths buffer satsfes equalty condton of (4.2). Thus we get probablty densty functon H H = M α f H (y, p, M, NH) =1 M y M y 1 = f α (x, p, NH) f α (x, p, NH) =1 x=0 =1 x=0 In ths way, we can get the TB n ths crcumstance: TB = 1 + H / L =1 l =1 H = 1 + lm L L = 1 + E(H ) / NH = 1 + yf H (y, p, M, NH) y=0 when L (4.6) (4.7) The (4.7) s also the lmt of TB wth FNC-ARQ scheme.
5 When the nteracton between the buffer and buffer are taen nto consderaton, some lost pacets n ths buffer may be recovered by the retransmsson pacets of the last buffer. After removng the recovery pacets due to ( 1)-th buffer, the number of lost pacets n th buffer can be rewrtten as β. As the example of DNC-ARQ n secton 3, we can wrte the β as: β = α ε (β 1 β 1 ) (4.8) where ε s a random varable from 0 to 1, β 1 (β 1 ). And then we have: TB DN C = 1 + (β ) / L =1 ( (β = 1 + E ) ) when L NH ( (α 1 + E ) ) = 1 + E( ) H NH HN Theorem 2. The low bound of TB wth DNC-ARQ s: where H f H (y, p, M, L) = TB DN C 1 + H / NH (4.9) Proof: Accordng to (4.8), the TB wth DNC-ARQ would reach mnmum f ε = 1, we have: So, β β β 1 β α (β 1 = α 1. (β 2 β 1 β 2 α 2 (β 1 β 1 ) β 1 α 1 ) β 2 ) can be rewrtten as: { τ=1 ατ 1 τ=1 βτ,wth 2 α,wth = 1 Here α τ f α(x, p, NH). We can get β from (4.10): β 1 (α 1 ) β 2 β 3. β Thus, we have: (α 2 + α 1 α) 1 ( α 3 + α 2 + α 1 (α 2 + α 1 ) ) ( τ=1 (β ) = β 1 α τ α τ ) τ=1 (4.10) A A 1 (4.11) where A f α (x, p, NH), A 1 f α (x, p, ( 1)NH), and A 0 = 0. Comparng to the expresson of (4.4), and we can wrte: H A = M =1 A f H (y, p, M, NH) M y M y 1 = f α (x, p, NH) f α (x, p, NH) =1 x=0 Thus, we get the low bound of TB: TB DN C = = 1 + =1 ( (β ) / L =1 A =1 x=0 A 1 ) / L ( H A 1 + H A 2 H A H A l H A l 1 = 1 + H A l /L here, H A l f H (y, p, M, L) When L +, we have: )/ L TB DN C = 1 / (1 p ) when, L + (4.12) B. Decodng Delay Defnton 5. Maxmum Decodng Delay The mum decodng delay of a NC-ARQ scheme can be wrtten as D = L ( ) µ(j) υ(j) j=1 where, Φ[µ(j), j] = Φ[υ(j), j] = 1, and Φ[, j] = 0, for [1,..., υ(j)] or [1,..., µ(j)] In DNC-ARQ scheme, the mum decodng delay occurred n the followng crcumstance: a lost pacet couldnt fnd ts prefect match whle ts survve tme s smaller than NH. Supposng t need h retransmsson pacets to ensure all UE can recover ths lost pacet.so the MTD of DNC-ARQ scheme can be wrtten as D = NH + h We can regard the lost pacet as a buffer wth sze=1, thus E(D) = NH + E(h ), h f H (y, p, M, 1) (4.13) We consder mum decodng delay wth FNC-ARQ scheme n the same way. In FNC-ARQ scheme, networ codng matrx of can be dvded nto submatrxes. Soppusng that the buffer sze s N, the mum decodng delay wth FNC-ARQ scheme can be wrtten as D = N + h But here, h f H(y, p, M, N), we have E(D ) = NH + E(h ) (4.14) Due to E(h ) > E(h ) when N > 1, we can get E(D) < E(D ) when NH = N (4.15)
6 The nequalty above suggest that: the decodng delay of DNC- ARQ scheme s smaller (n the math expectaton mean) than the decodng delay of FNC-ARQ, when the mum survve tme of DNC-ARQ s equal to the buffer sze of FNC-ARQ. V. SIMULATION AND RESULTS In ths secton, we gve numercal results to verfy the theoretcal analyss for DNC-ARQ scheme. The upper and low bound can be calculated by formula (4.5) and (4.9) when the pacet loss probablty pe, the number of UEs M, and the mum survve tme HN are gven. To obtan TB of DNC- ARQ schemes wth the proposed MMA, we assume the BS multcasts 500,000 orgnal pacets. The Fg.4 and Fg.5 show the smulaton results and the theoretcal result of the transmsson bandwdths of DNC-ARQ schemes for the scenaro consstng of 10 UEs. The average pacet loss probablty s 10% n Fg.4 and 30% n Fg.5 The low bound of DNC-ARQ(the blue lne) s equal to 1/(1 p) = 1/0.9 = approxmatvely n Fg.4 and 1/(1 p) = 1/0.7 = 1.43 n Fg.5. The low bound of DNC-ARQ n both two fgure are almost horzontal. The DNC-ARQ scheme (the blac lne) s always under the low bound of FNC-ARQ (the red lne). The TB of both are decreasng and convergng to the blue lne wth ncreasng of HN. To compare the TB of our proposed technques and the correspondng theoretcal bound. Fg.6 and Fg.7 show the varance of thetb wth pacet loss probablty at M = 5 and M = 50 respectvely. In ths scenaro, all UEs have the same the pacet loss probablty at every slot n multcast, and the mum survve tme s 20. All of the three lne s ncreasng wth pe. And only the blue lne stay unchanged n two fgures. Comparng all result above, we get the followng conclusons: 1) The TB of DNC-ARQ scheme s always under the low bound of FNC-ARQ, and that testfy our theoretcal analyss: the low bound of FNC-ARQ s also the upper bound of DNC-ARQ. 2) Increasng of NH have sgnfcant mprovement for TB of DNC-ARQ, and when NH s gettng larger t s convergng to the low bound. 3) The low bound of DNC-ARQ only depend on pacet loss probablty of UEs, and t approxmates to 1/(1 p) when the number of orgnal pacets s suffcently large. 4) The performance of DNC-ARQ s affected by number of UEs when NH s not suffcently large. It s less effcency when UEs s gettng more. REFERENCES [1] Mchael Luby, Tago Gasba, Thomas Stochammer, and Mar Watson, Relable Multmeda Download Delvery n Cellular Broadcast Networs, IEEE Transacton on Broadcastng, VOL. 53, NO. 1, PP , MAR [2] Shu L.,Phlp S.Y., A hybrd ARQ scheme wth party retransmsson for error control of satellte channels, IEEE transactons on communcatons, Vol.30(7) PP.78 92, [3] Salm Y.El Rouayheb, Mohammad Asad R.Chaudhry, and Alex Sprntson, On the Mnmum Number of Transmssons n Sngle-Hop Wreless Codng Networs, IEEE Informaton Theory Worshop, Lae Tahoe, Calforna, PP , Sep [4] Dong Nguyen,Tuan Tran,Thnh Nguyen,and Bella Bose, Wreless Broadcastng Usng Networ Codng, IEEE Transactons On Vehcular Technology, Vol.58, NO.2, PP , FEB Fg. 4. Fg. 5. Fg. 6. Fg. 7. TB of DNC-ARQ wth M=10,pe=0.1 TB of DNC-ARQ wth M=10,pe=0.3 TB of DNC-ARQ wth M=5,N=20 TB of DNC-ARQ wth M=50,N=20 [5] Qang.Hu, Jun Zheng, An effcent pacet selecton algorthm for networ codng based multcast retransmsson n moble communcaton networs, IEEE ICCT, Jnan,Chna,2011. [6] Qurensh J., Chuan Heng Foh, Janfe Ca, An effcent networ codng based retransmsson algorthm for wreless multcast, IEEE 20th Internatonal Sympo, Toyo,Japan,2009. [7] Kaa Ch, Xaohong Jang, Susumu Horguch, Networ codngbased relable multcast n wreless networ,computer Networ, Vol.54, PP , [8] H.D.T. Nguyen, L. Tran, E. Hong, On transmsson effcency for wreless broadcast usng networ codng and fountan codes, IEEE Communcatons Letters, Vol.15(5), PP , [9] Smeh Sorour, S. Valaee, An Adaptve Networ Coded Retransmsson Scheme for Sngle-Hop Wreless Multcast Broadcast Servces, ACM Trans. On Networng, Vol.19, No.3, Jun [10] K. Ch, X. Jang, S. Horguch, Bloc-level pacet recovery wth networ codng for wreless relable multcast, Computer Networ, Vol.57 PP , 2010.
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 informationModule 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 informationCOMPARISON 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 informationLecture 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 informationLecture 4: November 17, Part 1 Single Buffer Management
Lecturer: Ad Rosén Algorthms for the anagement of Networs Fall 2003-2004 Lecture 4: November 7, 2003 Scrbe: Guy Grebla Part Sngle Buffer anagement In the prevous lecture we taled about the Combned Input
More informationChapter 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 informationEncoded packet-assisted Rescue Approach to Reliable Unicast in Wireless Networks
Encoded pacet-asssted Rescue Approach to Relable Uncast n Wreless Networs Zhheng Zhou, Lang Zhou, Xng Wang, Yuanquan Tan Natonal Key Laboratory of Communcaton Unversty of Electronc Scence and Tech of Chna
More informationEEE 241: Linear Systems
EEE : Lnear Systems Summary #: Backpropagaton BACKPROPAGATION The perceptron rule as well as the Wdrow Hoff learnng were desgned to tran sngle layer networks. They suffer from the same dsadvantage: they
More informationGrover s Algorithm + Quantum Zeno Effect + Vaidman
Grover s Algorthm + Quantum Zeno Effect + Vadman CS 294-2 Bomb 10/12/04 Fall 2004 Lecture 11 Grover s algorthm Recall that Grover s algorthm for searchng over a space of sze wors as follows: consder the
More informationThe 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 informationFoundations of Arithmetic
Foundatons of Arthmetc Notaton We shall denote the sum and product of numbers n the usual notaton as a 2 + a 2 + a 3 + + a = a, a 1 a 2 a 3 a = a The notaton a b means a dvdes b,.e. ac = b where c s an
More informationNetwork Coding Based Schemes for Imperfect Wireless Packet Retransmission Problems: A Divide and Conquer Approach
Wreless Pers Commun DOI 101007/s11277-010-0102-9 Network Codng Based Schemes for Imperfect Wreless Packet Retransmsson Problems: A Dvde and Conquer Approach Zhenguo Gao Me Yang Janpng Wang Klara Nahrstedt
More informationPower law and dimension of the maximum value for belief distribution with the max Deng entropy
Power law and dmenson of the maxmum value for belef dstrbuton wth the max Deng entropy Bngy Kang a, a College of Informaton Engneerng, Northwest A&F Unversty, Yanglng, Shaanx, 712100, Chna. Abstract Deng
More informationEstimation: 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 informationA PROBABILITY-DRIVEN SEARCH ALGORITHM FOR SOLVING MULTI-OBJECTIVE OPTIMIZATION PROBLEMS
HCMC Unversty of Pedagogy Thong Nguyen Huu et al. A PROBABILITY-DRIVEN SEARCH ALGORITHM FOR SOLVING MULTI-OBJECTIVE OPTIMIZATION PROBLEMS Thong Nguyen Huu and Hao Tran Van Department of mathematcs-nformaton,
More informationSystem in Weibull Distribution
Internatonal Matheatcal Foru 4 9 no. 9 94-95 Relablty Equvalence Factors of a Seres-Parallel Syste n Webull Dstrbuton M. A. El-Dacese Matheatcs Departent Faculty of Scence Tanta Unversty Tanta Egypt eldacese@yahoo.co
More informationProblem 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 informationResource 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 information3.1 Expectation of Functions of Several Random Variables. )' be a k-dimensional discrete or continuous random vector, with joint PMF p (, E X E X1 E X
Statstcs 1: Probablty Theory II 37 3 EPECTATION OF SEVERAL RANDOM VARIABLES As n Probablty Theory I, the nterest n most stuatons les not on the actual dstrbuton of a random vector, but rather on a number
More informationANSWERS. 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 informationCollege 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 informationOn the correction of the h-index for career length
1 On the correcton of the h-ndex for career length by L. Egghe Unverstet Hasselt (UHasselt), Campus Depenbeek, Agoralaan, B-3590 Depenbeek, Belgum 1 and Unverstet Antwerpen (UA), IBW, Stadscampus, Venusstraat
More informationA 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 informationExpected Value and Variance
MATH 38 Expected Value and Varance Dr. Neal, WKU We now shall dscuss how to fnd the average and standard devaton of a random varable X. Expected Value Defnton. The expected value (or average value, or
More informationEcon107 Applied Econometrics Topic 3: Classical Model (Studenmund, Chapter 4)
I. Classcal Assumptons Econ7 Appled Econometrcs Topc 3: Classcal Model (Studenmund, Chapter 4) We have defned OLS and studed some algebrac propertes of OLS. In ths topc we wll study statstcal propertes
More informationMin Cut, Fast Cut, Polynomial Identities
Randomzed Algorthms, Summer 016 Mn Cut, Fast Cut, Polynomal Identtes Instructor: Thomas Kesselhem and Kurt Mehlhorn 1 Mn Cuts n Graphs Lecture (5 pages) Throughout ths secton, G = (V, E) s a mult-graph.
More informationCHAPTER 5 NUMERICAL EVALUATION OF DYNAMIC RESPONSE
CHAPTER 5 NUMERICAL EVALUATION OF DYNAMIC RESPONSE Analytcal soluton s usually not possble when exctaton vares arbtrarly wth tme or f the system s nonlnear. Such problems can be solved by numercal tmesteppng
More informationMMA 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 informationLow Complexity Soft-Input Soft-Output Hamming Decoder
Low Complexty Soft-Input Soft-Output Hammng Der Benjamn Müller, Martn Holters, Udo Zölzer Helmut Schmdt Unversty Unversty of the Federal Armed Forces Department of Sgnal Processng and Communcatons Holstenhofweg
More informationLectures - Week 4 Matrix norms, Conditioning, Vector Spaces, Linear Independence, Spanning sets and Basis, Null space and Range of a Matrix
Lectures - Week 4 Matrx norms, Condtonng, Vector Spaces, Lnear Independence, Spannng sets and Bass, Null space and Range of a Matrx Matrx Norms Now we turn to assocatng a number to each matrx. We could
More informationError Probability for M Signals
Chapter 3 rror Probablty for M Sgnals In ths chapter we dscuss the error probablty n decdng whch of M sgnals was transmtted over an arbtrary channel. We assume the sgnals are represented by a set of orthonormal
More informationAppendix B. The Finite Difference Scheme
140 APPENDIXES Appendx B. The Fnte Dfference Scheme In ths appendx we present numercal technques whch are used to approxmate solutons of system 3.1 3.3. A comprehensve treatment of theoretcal and mplementaton
More informationKernel 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 informationDUE: WEDS FEB 21ST 2018
HOMEWORK # 1: FINITE DIFFERENCES IN ONE DIMENSION DUE: WEDS FEB 21ST 2018 1. Theory Beam bendng s a classcal engneerng analyss. The tradtonal soluton technque makes smplfyng assumptons such as a constant
More informationStanford University CS359G: Graph Partitioning and Expanders Handout 4 Luca Trevisan January 13, 2011
Stanford Unversty CS359G: Graph Parttonng and Expanders Handout 4 Luca Trevsan January 3, 0 Lecture 4 In whch we prove the dffcult drecton of Cheeger s nequalty. As n the past lectures, consder an undrected
More informationMarkov Chain Monte Carlo Lecture 6
where (x 1,..., x N ) X N, N s called the populaton sze, f(x) f (x) for at least one {1, 2,..., N}, and those dfferent from f(x) are called the tral dstrbutons n terms of mportance samplng. Dfferent ways
More informationNUMERICAL DIFFERENTIATION
NUMERICAL DIFFERENTIATION 1 Introducton Dfferentaton s a method to compute the rate at whch a dependent output y changes wth respect to the change n the ndependent nput x. Ths rate of change s called the
More informationDr. Shalabh Department of Mathematics and Statistics Indian Institute of Technology Kanpur
Analyss of Varance and Desgn of Exerments-I MODULE III LECTURE - 2 EXPERIMENTAL DESIGN MODELS Dr. Shalabh Deartment of Mathematcs and Statstcs Indan Insttute of Technology Kanur 2 We consder the models
More informationA new construction of 3-separable matrices via an improved decoding of Macula s construction
Dscrete Optmzaton 5 008 700 704 Contents lsts avalable at ScenceDrect Dscrete Optmzaton journal homepage: wwwelsevercom/locate/dsopt A new constructon of 3-separable matrces va an mproved decodng of Macula
More informationA Robust Method for Calculating the Correlation Coefficient
A Robust Method for Calculatng the Correlaton Coeffcent E.B. Nven and C. V. Deutsch Relatonshps between prmary and secondary data are frequently quantfed usng the correlaton coeffcent; however, the tradtonal
More informationOutline. Communication. Bellman Ford Algorithm. Bellman Ford Example. Bellman Ford Shortest Path [1]
DYNAMIC SHORTEST PATH SEARCH AND SYNCHRONIZED TASK SWITCHING Jay Wagenpfel, Adran Trachte 2 Outlne Shortest Communcaton Path Searchng Bellmann Ford algorthm Algorthm for dynamc case Modfcatons to our algorthm
More informationLINEAR REGRESSION ANALYSIS. MODULE IX Lecture Multicollinearity
LINEAR REGRESSION ANALYSIS MODULE IX Lecture - 30 Multcollnearty Dr. Shalabh Department of Mathematcs and Statstcs Indan Insttute of Technology Kanpur 2 Remedes for multcollnearty Varous technques have
More informationLecture 14 (03/27/18). Channels. Decoding. Preview of the Capacity Theorem.
Lecture 14 (03/27/18). Channels. Decodng. Prevew of the Capacty Theorem. A. Barg The concept of a communcaton channel n nformaton theory s an abstracton for transmttng dgtal (and analog) nformaton from
More informationStructure and Drive Paul A. Jensen Copyright July 20, 2003
Structure and Drve Paul A. Jensen Copyrght July 20, 2003 A system s made up of several operatons wth flow passng between them. The structure of the system descrbes the flow paths from nputs to outputs.
More informationAnnexes. EC.1. Cycle-base move illustration. EC.2. Problem Instances
ec Annexes Ths Annex frst llustrates a cycle-based move n the dynamc-block generaton tabu search. It then dsplays the characterstcs of the nstance sets, followed by detaled results of the parametercalbraton
More informationThe optimal delay of the second test is therefore approximately 210 hours earlier than =2.
THE IEC 61508 FORMULAS 223 The optmal delay of the second test s therefore approxmately 210 hours earler than =2. 8.4 The IEC 61508 Formulas IEC 61508-6 provdes approxmaton formulas for the PF for smple
More informationChapter 7 Channel Capacity and Coding
Wreless Informaton Transmsson System Lab. Chapter 7 Channel Capacty and Codng Insttute of Communcatons Engneerng atonal Sun Yat-sen Unversty Contents 7. Channel models and channel capacty 7.. Channel models
More informationNegative Binomial Regression
STATGRAPHICS Rev. 9/16/2013 Negatve Bnomal Regresson Summary... 1 Data Input... 3 Statstcal Model... 3 Analyss Summary... 4 Analyss Optons... 7 Plot of Ftted Model... 8 Observed Versus Predcted... 10 Predctons...
More informationThe 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 informationRate-Memory Trade-off for the Two-User Broadcast Caching Network with Correlated Sources
Rate-Memory Trade-off for the Two-User Broadcast Cachng Network wth Correlated Sources Parsa Hassanzadeh, Antona M. Tulno, Jame Llorca, Elza Erkp arxv:705.0466v [cs.it] May 07 Abstract Ths paper studes
More informationn 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 informationThe 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 information1 GSW Iterative Techniques for y = Ax
1 for y = A I m gong to cheat here. here are a lot of teratve technques that can be used to solve the general case of a set of smultaneous equatons (wrtten n the matr form as y = A), but ths chapter sn
More informationMASSACHUSETTS INSTITUTE OF TECHNOLOGY 6.265/15.070J Fall 2013 Lecture 12 10/21/2013. Martingale Concentration Inequalities and Applications
MASSACHUSETTS INSTITUTE OF TECHNOLOGY 6.65/15.070J Fall 013 Lecture 1 10/1/013 Martngale Concentraton Inequaltes and Applcatons Content. 1. Exponental concentraton for martngales wth bounded ncrements.
More informationCOS 521: Advanced Algorithms Game Theory and Linear Programming
COS 521: Advanced Algorthms Game Theory and Lnear Programmng Moses Charkar February 27, 2013 In these notes, we ntroduce some basc concepts n game theory and lnear programmng (LP). We show a connecton
More informationTransfer Functions. Convenient representation of a linear, dynamic model. A transfer function (TF) relates one input and one output: ( ) system
Transfer Functons Convenent representaton of a lnear, dynamc model. A transfer functon (TF) relates one nput and one output: x t X s y t system Y s The followng termnology s used: x y nput output forcng
More informationLecture 3: Shannon s Theorem
CSE 533: Error-Correctng Codes (Autumn 006 Lecture 3: Shannon s Theorem October 9, 006 Lecturer: Venkatesan Guruswam Scrbe: Wdad Machmouch 1 Communcaton Model The communcaton model we are usng conssts
More informationThe Multiple Classical Linear Regression Model (CLRM): Specification and Assumptions. 1. Introduction
ECONOMICS 5* -- NOTE (Summary) ECON 5* -- NOTE The Multple Classcal Lnear Regresson Model (CLRM): Specfcaton and Assumptons. Introducton CLRM stands for the Classcal Lnear Regresson Model. The CLRM s also
More informationCONTRAST ENHANCEMENT FOR MIMIMUM MEAN BRIGHTNESS ERROR FROM HISTOGRAM PARTITIONING INTRODUCTION
CONTRAST ENHANCEMENT FOR MIMIMUM MEAN BRIGHTNESS ERROR FROM HISTOGRAM PARTITIONING N. Phanthuna 1,2, F. Cheevasuvt 2 and S. Chtwong 2 1 Department of Electrcal Engneerng, Faculty of Engneerng Rajamangala
More informationHongyi 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 informationErrors 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 informationSimulated Power of the Discrete Cramér-von Mises Goodness-of-Fit Tests
Smulated of the Cramér-von Mses Goodness-of-Ft Tests Steele, M., Chaselng, J. and 3 Hurst, C. School of Mathematcal and Physcal Scences, James Cook Unversty, Australan School of Envronmental Studes, Grffth
More informationUsing T.O.M to Estimate Parameter of distributions that have not Single Exponential Family
IOSR Journal of Mathematcs IOSR-JM) ISSN: 2278-5728. Volume 3, Issue 3 Sep-Oct. 202), PP 44-48 www.osrjournals.org Usng T.O.M to Estmate Parameter of dstrbutons that have not Sngle Exponental Famly Jubran
More informationDesign 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 informationAn Upper Bound on SINR Threshold for Call Admission Control in Multiple-Class CDMA Systems with Imperfect Power-Control
An Upper Bound on SINR Threshold for Call Admsson Control n Multple-Class CDMA Systems wth Imperfect ower-control Mahmoud El-Sayes MacDonald, Dettwler and Assocates td. (MDA) Toronto, Canada melsayes@hotmal.com
More informationLecture 4: Constant Time SVD Approximation
Spectral Algorthms and Representatons eb. 17, Mar. 3 and 8, 005 Lecture 4: Constant Tme SVD Approxmaton Lecturer: Santosh Vempala Scrbe: Jangzhuo Chen Ths topc conssts of three lectures 0/17, 03/03, 03/08),
More informationCalculation of time complexity (3%)
Problem 1. (30%) Calculaton of tme complexty (3%) Gven n ctes, usng exhaust search to see every result takes O(n!). Calculaton of tme needed to solve the problem (2%) 40 ctes:40! dfferent tours 40 add
More informationMore metrics on cartesian products
More metrcs on cartesan products If (X, d ) are metrc spaces for 1 n, then n Secton II4 of the lecture notes we defned three metrcs on X whose underlyng topologes are the product topology The purpose of
More informationAn Improved multiple fractal algorithm
Advanced Scence and Technology Letters Vol.31 (MulGraB 213), pp.184-188 http://dx.do.org/1.1427/astl.213.31.41 An Improved multple fractal algorthm Yun Ln, Xaochu Xu, Jnfeng Pang College of Informaton
More information4 Analysis of Variance (ANOVA) 5 ANOVA. 5.1 Introduction. 5.2 Fixed Effects ANOVA
4 Analyss of Varance (ANOVA) 5 ANOVA 51 Introducton ANOVA ANOVA s a way to estmate and test the means of multple populatons We wll start wth one-way ANOVA If the populatons ncluded n the study are selected
More informationTurbulence classification of load data by the frequency and severity of wind gusts. Oscar Moñux, DEWI GmbH Kevin Bleibler, DEWI GmbH
Turbulence classfcaton of load data by the frequency and severty of wnd gusts Introducton Oscar Moñux, DEWI GmbH Kevn Blebler, DEWI GmbH Durng the wnd turbne developng process, one of the most mportant
More informationA Network Intrusion Detection Method Based on Improved K-means Algorithm
Advanced Scence and Technology Letters, pp.429-433 http://dx.do.org/10.14257/astl.2014.53.89 A Network Intruson Detecton Method Based on Improved K-means Algorthm Meng Gao 1,1, Nhong Wang 1, 1 Informaton
More informationDr. Shalabh Department of Mathematics and Statistics Indian Institute of Technology Kanpur
Analyss of Varance and Desgn of Experment-I MODULE VII LECTURE - 3 ANALYSIS OF COVARIANCE Dr Shalabh Department of Mathematcs and Statstcs Indan Insttute of Technology Kanpur Any scentfc experment s performed
More informationLOW 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 informationAnalysis of Discrete Time Queues (Section 4.6)
Analyss of Dscrete Tme Queues (Secton 4.6) Copyrght 2002, Sanjay K. Bose Tme axs dvded nto slots slot slot boundares Arrvals can only occur at slot boundares Servce to a job can only start at a slot boundary
More informationModule 9. Lecture 6. Duality in Assignment Problems
Module 9 1 Lecture 6 Dualty n Assgnment Problems In ths lecture we attempt to answer few other mportant questons posed n earler lecture for (AP) and see how some of them can be explaned through the concept
More informationThe 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 informationPower 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 informationChapter - 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 informationComparison of Regression Lines
STATGRAPHICS Rev. 9/13/2013 Comparson of Regresson Lnes Summary... 1 Data Input... 3 Analyss Summary... 4 Plot of Ftted Model... 6 Condtonal Sums of Squares... 6 Analyss Optons... 7 Forecasts... 8 Confdence
More informationGeneralized 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 informationA 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 informationAppendix B. Criterion of Riemann-Stieltjes Integrability
Appendx B. Crteron of Remann-Steltes Integrablty Ths note s complementary to [R, Ch. 6] and [T, Sec. 3.5]. The man result of ths note s Theorem B.3, whch provdes the necessary and suffcent condtons for
More informationSuppose that there s a measured wndow of data fff k () ; :::; ff k g of a sze w, measured dscretely wth varable dscretzaton step. It s convenent to pl
RECURSIVE SPLINE INTERPOLATION METHOD FOR REAL TIME ENGINE CONTROL APPLICATIONS A. Stotsky Volvo Car Corporaton Engne Desgn and Development Dept. 97542, HA1N, SE- 405 31 Gothenburg Sweden. Emal: astotsky@volvocars.com
More informationA new Approach for Solving Linear Ordinary Differential Equations
, ISSN 974-57X (Onlne), ISSN 974-5718 (Prnt), Vol. ; Issue No. 1; Year 14, Copyrght 13-14 by CESER PUBLICATIONS A new Approach for Solvng Lnear Ordnary Dfferental Equatons Fawz Abdelwahd Department of
More informationχ x B E (c) Figure 2.1.1: (a) a material particle in a body, (b) a place in space, (c) a configuration of the body
Secton.. Moton.. The Materal Body and Moton hyscal materals n the real world are modeled usng an abstract mathematcal entty called a body. Ths body conssts of an nfnte number of materal partcles. Shown
More informationSociété de Calcul Mathématique SA
Socété de Calcul Mathématque SA Outls d'ade à la décson Tools for decson help Probablstc Studes: Normalzng the Hstograms Bernard Beauzamy December, 202 I. General constructon of the hstogram Any probablstc
More informationLecture Randomized Load Balancing strategies and their analysis. Probability concepts include, counting, the union bound, and Chernoff bounds.
U.C. Berkeley CS273: Parallel and Dstrbuted Theory Lecture 1 Professor Satsh Rao August 26, 2010 Lecturer: Satsh Rao Last revsed September 2, 2010 Lecture 1 1 Course Outlne We wll cover a samplng of the
More informationfind (x): given element x, return the canonical element of the set containing x;
COS 43 Sprng, 009 Dsjont Set Unon Problem: Mantan a collecton of dsjont sets. Two operatons: fnd the set contanng a gven element; unte two sets nto one (destructvely). Approach: Canoncal element method:
More informationTHE ARIMOTO-BLAHUT ALGORITHM FOR COMPUTATION OF CHANNEL CAPACITY. William A. Pearlman. References: S. Arimoto - IEEE Trans. Inform. Thy., Jan.
THE ARIMOTO-BLAHUT ALGORITHM FOR COMPUTATION OF CHANNEL CAPACITY Wllam A. Pearlman 2002 References: S. Armoto - IEEE Trans. Inform. Thy., Jan. 1972 R. Blahut - IEEE Trans. Inform. Thy., July 1972 Recall
More informationErratum: A Generalized Path Integral Control Approach to Reinforcement Learning
Journal of Machne Learnng Research 00-9 Submtted /0; Publshed 7/ Erratum: A Generalzed Path Integral Control Approach to Renforcement Learnng Evangelos ATheodorou Jonas Buchl Stefan Schaal Department of
More informationOpen Systems: Chemical Potential and Partial Molar Quantities Chemical Potential
Open Systems: Chemcal Potental and Partal Molar Quanttes Chemcal Potental For closed systems, we have derved the followng relatonshps: du = TdS pdv dh = TdS + Vdp da = SdT pdv dg = VdP SdT For open systems,
More informationDifference Equations
Dfference Equatons c Jan Vrbk 1 Bascs Suppose a sequence of numbers, say a 0,a 1,a,a 3,... s defned by a certan general relatonshp between, say, three consecutve values of the sequence, e.g. a + +3a +1
More informationRefined Coding Bounds for Network Error Correction
Refned Codng Bounds for Network Error Correcton Shenghao Yang Department of Informaton Engneerng The Chnese Unversty of Hong Kong Shatn, N.T., Hong Kong shyang5@e.cuhk.edu.hk Raymond W. Yeung Department
More information18.1 Introduction and Recap
CS787: Advanced Algorthms Scrbe: Pryananda Shenoy and Shjn Kong Lecturer: Shuch Chawla Topc: Streamng Algorthmscontnued) Date: 0/26/2007 We contnue talng about streamng algorthms n ths lecture, ncludng
More informationHow Strong Are Weak Patents? Joseph Farrell and Carl Shapiro. Supplementary Material Licensing Probabilistic Patents to Cournot Oligopolists *
How Strong Are Weak Patents? Joseph Farrell and Carl Shapro Supplementary Materal Lcensng Probablstc Patents to Cournot Olgopolsts * September 007 We study here the specal case n whch downstream competton
More informationComments on a secure dynamic ID-based remote user authentication scheme for multiserver environment using smart cards
Comments on a secure dynamc ID-based remote user authentcaton scheme for multserver envronment usng smart cards Debao He chool of Mathematcs tatstcs Wuhan nversty Wuhan People s Republc of Chna Emal: hedebao@63com
More informationOn 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 informationEGR 544 Communication Theory
EGR 544 Communcaton Theory. Informaton Sources Z. Alyazcoglu Electrcal and Computer Engneerng Department Cal Poly Pomona Introducton Informaton Source x n Informaton sources Analog sources Dscrete sources
More informationVQ 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