Downlink Throughput Maximization: TDMA versus CDMA

Size: px
Start display at page:

Download "Downlink Throughput Maximization: TDMA versus CDMA"

Transcription

1 00 Conference on Informaton Scences and Systems, The Johns Hopkns Unversty, March, 00 Downlnk Throughput Maxmzaton: TDMA versus CDMA Changyoon Oh and Ayln Yener Department of Electrcal Engneerng The Pennsylvana State Unversty Unversty Park, PA 0 e-mal: changyoon@psu.edu yener@ee.psu.edu Abstract We consder the downlnk throughput maxmzaton problem for nterference lmted multuser systems. Our goal s to characterze the optmum base staton transmsson polcy (TDMA or CDMA), and fnd the correspondng power allocaton. We model the nterference by the ad of the orthogonalty factor and determne the transmsson strategy for a range of the values of the orthogonalty factor, and users channel gans, so as to maxmze the system throughput, subject to a total power constrant. Although the resultng optmzaton problem may turn out to be non-convex dependng on the value of the orthogonalty factor, we show that valuable observatons regardng the optmum soluton can be obtaned based on the propertes of the problem. In partcular, we propose an exact and a near-exact algorthm to determne whether TDMA s the optmum strategy, or smultaneous users (CDMA) should be transmtted to. I. Introducton Wreless servces are becomng more data centrc and often requre hgher data rate for downlnk communcatons [, ]. Data servces typcally requre hgh relablty but are delay tolerant. The delay tolerance can be exploted by effectvely schedulng transmssons to users n favorable channel condtons to ncrease the overall system throughput []. In ths paper, we consder the power allocaton problem for the downlnk of delay tolerant CDMA systems, by takng the channels of the users nto account. For the CDMA downlnk, typcally, orthogonal codes are used to communcate to each user []. However, due to multpath fadng, orthogonalty between codes s not preserved at the moble sde. In such cases, a parameter defned as the orthogonalty factor s often used to desgnate the level of multpath fadng, representng the degree of nterference from other users []. We focus on ths scenaro, where the orthogonalty between the codes s not mantaned. Specfcally, our am s to fnd the optmum transmsson polcy,.e., whch user(s) wll be transmtted to and the power allocaton n each tme slot so as to maxmze the total system utlty, subject to a total power constrant at the base staton. The total system utlty s defned smply as the sum of the ndvdual utltes. The ndvdual utlty we consder n ths paper s the transmsson rate to a user, whch s an ncreasng functon of the user s transmt power. Utlty based power control for data servces has prevously been consdered n [, ], for the uplnk, where the qualty of servce (QoS) of each user, e.g. sgnal to nterference rato (SIR), s defned as the utlty. In references [, 9], one-at-a-tme transmsson for downlnk throughput maxmzaton s consdered for delay tolerant servce, and the optmalty of ths soluton s shown. All avalable power from the base staton s transmtted for the sake of ncreased throughput, and no compensaton for the ntracell nterference s needed [, 9]. To overcome the possble unfarness that may result from the one-at-a-tme transmsson strategy, reference [0] mposes a mnmum servce rate requrement for each user, and consders the throughput maxmzaton problem for downlnk multrate CDMA system where the multrate s acheved by ether multcode or orthogonal varable spreadng factor (OVSF). The optmalty of greedy power allocaton s shown n ths scenaro [0]. Ths polcy s ntutvely pleasng when the user rate and power have a lnear relatonshp such as n a CDMA system wth varable processng gan or multcodes. Reference [0] also provdes numercal results whch show the effect of orthogonalty factor to the system performance. Agrawal et al. n [] consder the optmzaton of the sum of the ndvdual weghted throughput values n the downlnk of a multrate CDMA system wth orthogonal codes. It s assumed that the orthogonalty s preserved at the recever sde. The resultng problem s a convex program, whch for the specal case of the equal weght scenaro,.e., throughput maxmzaton, produces one-at-a-tme transmsson as the optmum polcy []. In contrast wth [], we consder the downlnk of an nterference lmted CDMA system and determne the optmum transmsson strategy. The key observaton s that the optmum polcy s a functon of the orthogonalty factor. Specfcally, gven the channel realzaton of the users, we determne whether the TDMA-mode,.e., the base staton transmts to one user at a tme, or the CDMA-mode,.e., smultaneous transmssons, should be chosen. If the CDMA-mode s optmum, that s, f multple nterferng users should be served by the base staton smultaneously, then the correspondng power allocaton s also found. The paper s organzed as follows. Secton II descrbes the system model as well as the problem formulaton and the performance metrc. Optmum power allocaton s consdered n secton III. Secton IV provdes numercal examples that support our analyss, and Secton V summarzes the results and concludes the paper. II. System Model and Problem Formulaton We consder a sngle cell CDMA system wth K users. The base staton has a maxmum power budget, P T. Ths power budget s assgned to each user such that p PT, where p s the allocated power for user. Fgure depcts the model of the CDMA downlnk. An orthogonal code s assgned to each user and s used to modulate the sgnal transmtted to that user. However, due to multpath fadng, orthogonalty between the codes (users) s lost at the recever sde. The

2 User User User K g g g K p p p K Base Staton Orthogonal codes code code code K where we assume that g > g >... > g K, such that user wth the lower ndex has a hgher channel gan. We note that the actual outcome of the above optmzaton problem s power vector. The optmum transmsson strategy s mplctly ncluded n the optmum power vector n that the users that belong to the subset of users that the base staton transmts to, end up wth non-zero power, and the rest wth zero power. If the TDMA-mode turns out to be optmum, the power allocaton s such that transmsson power to only one user s postve. If the CDMA-mode turns out to be optmum, the power allocaton s such that transmtted power values to multple users are postve. We frst note that any power vector p wth p < P T can not be the optmum power vector. Defne a power vector p wth p = βp (β > ), =,, K such that βp = P T. It s easy to see that p ncreases all ndvdual utltes and hence the system utlty as compared to p. Thus, () can be rewrtten as: Fg. : Downlnk System model degree of nonorthogonalty s descrbed by the orthogonalty factor, (0 ) []. Under the assumpton of ndependent dentcal channels for all users, t s reasonable to work wth the same average orthogonalty factor for all users. We assume that each user employs a matched flter recever. The sgnal to nterference rato at the recever of user s expressed as p g SIR = j pjg + I () where g denotes channel gan of user. I denotes the sum of the thermal nose and the ntercell nterference. We defne the utlty of user as follows: p g U (p) = log( + k ) () pjg + I j where log denotes the natural logarthm. We note that the ndvdual utlty s a functon of the power vector p = [p, p,..., p K]. The utlty s an achevable rate for user, specfcally by treatng the nterference as nose [9]. It s assumed that adaptve modulaton s employed to enable the rate determned for each user []. The factor k captures the product of the adaptve modulaton ndex and the fxed processng gan. Ths resembles the structure of HSDPA where multrate s acheved by multcode and adaptve modulaton. We should also note that, n a practcal system employng M- ary modulaton, the transmsson rate (utlty) s a dscrete quantty. For smplcty of analyss, and smlar to prevous work [, ], we wll assume contnuous rate values. We wll examne the effect of ths assumpton n the numercal results. The problem we consder s to determne the optmum allocaton of the total transmt power from the base staton to the users so as to maxmze the overall system utlty,.e., sum of ndvdual utltes. Formally, the optmzaton problem s formulated as: log( + k p g max p s.t. j p jg +I ) () p PT, p 0, =,, K () max p s.t. log( + k p ) () (p T p )+ g I p = PT, p 0, =,, K () Observe that the utlty of user s a functon of the power for user only,.e., U (p) = U (p ). Also note that, although the utlty s concave n SIR, t need not be concave n power. Therefore, the optmzaton problem s n general not a convex program. Fgure emphaszes ths pont by plottng the ndvdual utltes for users wth dfferent channel gans. III. Optmum Transmt Strategy and Power Allocaton We examne the throughput maxmzaton problem n terms of the ndvdual utlty and the system utlty, whch we term the user centrc approach, and the system-wde approach, respectvely. We frst consder the user centrc approach, and we observe the behavor of the ndvdual utlty n power by utlzng the dervatve of the ndvdual utlty functon. Our observatons motvate us to consder the system-wde approach, n whch we nvestgate the system utlty n terms of the best user s power. The system-wde approach s useful to characterze the system utlty when the utltes of all users are not concave n power. Fgure depcts such a case. III.A The User Centrc Approach Consder the utlty functon of user and ts frst and second dervatves n terms of p, the power of user. The frst dervatve of the utlty functon s postve,.e., U (p ) s an ncreasng functon of p. We are nterested n whether the utlty U (p ) s an ncreasng concave or convex because ths plays a crucal role n characterzng the behavor of the utlty n power. Whether U (p ) s concave ( U (p ) < 0 ) or convex p ( U (p ) p > 0) depends on the sgn(a ) of the numerator of U (p ). A p s gven by A = ( k)((p T p ) + I g ) + ((p T p ) + I g + kp ). By lettng I t = (p T p ) + I g, we have A = ( k)i t + (I t + kp ). Thus, we need to have p I t > k ()

3 . λ mn 0 9. λ max p s wthn concave regon p n p s wthn convex regon λ max gan λ mn Utlty. gan Utlty gan. λ max λ max gan gan gan 0. gan gan Transmt power[watts] Transmt power[watts] Fg. : gan > gan > gan > gan. An example where ndvdual utlty functons for all users are concave. Fg. : gan > gan > gan > gan. An example where ndvdual utlty functons for some users are not concave. for U (p ) to be an ncreasng convex functon of p and p I t < k for U (p ) to be an ncreasng concave functon of p. Clearly, by observng U (p ) p p =P T, we can dentfy the behavor of functon U (p ). When U (p ) p p =P T < 0, U (p ) s an ncreasng concave functon n 0 p P T as n Fgure. When p =P T > 0, U (p ) s an ncreasng concave functon U (p ) p n 0 p p n, whle t s ncreasng convex functon n p n p P T, where p n denotes the nflecton pont n U (p ) as n U (p ) shown n Fgure. Fgure shows the ndvdual utltes of dfferent channel gans. In ths example, the ndvdual utltes of all users are concave n power. It s clear that when the ndvdual utlty s concave n power, among multple users, rather than allocatng the entre power P T to one user, sharng the power P T wll yeld a hgher total utlty. In ths case, the optmzaton problem n () s a convex program. However, ths s no longer the case when the ndvdual utltes of some users are not concave, as depcted n Fgure. In ths case, we need to have a closer look at the components that contrbute to the system utlty. The followng termnology s used extensvely n the sequel. U (p ) s convex f U (p ) > 0 n 0 p p P T U (p ) s concave/convex f U (p ) p p =0< 0 and U (p ) p p =P T > 0 U (p ) s concave f U (p ) p p =P T < 0 p s sad to be wthn the concave regon of U (p ), f U (p ) p p =p < 0 p s sad to be wthn the convex regon of U (p ), f U (p ) p p =p > 0 () Fgure depcts two possble power regons of user (concave/convex user). Followng from the above, we defne λ mn = U (p ) p p =p whenu(p)s concave/convex, (9) λ mn = λ max = and, and U (p ) p p =p = 0 ; U (p ) p p =pt whenu(p)s concave (0) U (p ) p p =0 () p (λ) = arg p ( U (p ) p p =p (λ) = λ) whenλ < λ max() = 0 whenλ > λ max() where λ mn mples the mnmum dervatve value of utlty of user, whle λ max denotes the maxmum dervatve value of utlty of user. Fgures and depct λ mn and λ max. Gven λ, power of user p (λ) s obtaned by optmzng the ndvdual utlty U (p ) over p. The optmum polcy s a functon of the orthogonalty factor as well as users channel gans n general. Below we state the propostons that lead to characterzng the optmum polces. The proofs of all the propostons n ths paper can be found n []. Our frst observaton states that when the orthogonalty factor s larger than a certan threshold, TDMA-mode s always optmum, regardless of users channel gans: Proposton III.: If k, TDMA-mode wth p = PT yelds the optmum system utlty. Whle, t s clear that the TDMA-mode s optmum when k, and no further power allocaton s necessary, when k, we need to characterze the optmum polcy whch conssts of the transmsson strategy,.e., TDMA-mode or CDMA-mode, and the correspondng power allocaton. Frst, we observe the followng. Proposton III.: The optmum power allocaton p = (p, p,..., p K) s such that p > p >... > p K when p > 0.

4 When the ndvdual utltes of multple users are concave/convex as n Fgure, we have the followng observaton. Proposton III.: Suppose p s the optmum power allocaton. Then, at most one user s power (the user wth the best channel gan) s wthn the convex regon of the utlty functon. From the precedng observatons, we conclude that the optmum power allocaton p = [p, p,..., p K] s one of two cases n terms of the best user,.e., p can be ether wthn the concave regon or wthn the convex regon of U (p ). Ths observaton motvates our approach to focus on the the system utlty n terms of the best user. III.B System-Wde Approach We defne the system utlty n terms of the best user,.e., p : J(p ) = U (p ) + Z(p ) () Z(p ) = max[ j= Uj(pj)] j= p j=p T p. () We note that J(p ) expends all avalable power, P T, p for user and P T p for the rest of the users. Our am s to fnd the maxmzer of J(p ) over 0 p P T. Proposton guarantees that U j(p j) (j ) s always concave at optmum power allocaton. Hence, Z(p ) can be maxmzed wth the power constrant pj(λ) = PT p where λ s j= the optmum Lagrange multpler. We note that no power s assgned to the user when λ max < λ. Dependng on user s channel condton, U (p ) can be ether concave or concave/convex. When U (p ) s concave, the resultng optmzaton problem s convex and the optmum power allocaton s such that pj(λ) = PT. If U(p) s concave/convex j= as n Fgure, the optmum power allocaton s one of three possbltes. Observaton III.: If U (p ) s concave/convex as n Fgure, the optmum power allocaton s one of three local optmum solutons: () p s wthn the concave regon wth 0 p < p (λ mn), () p s wthn the convex regon wth p (λ mn) < p < P T, or () p = P T. Transmsson strategy for the frst two cases s CDMA, whle the thrd case (p = P T) s TDMA. Gven the fact that when U (p ) s concave/convex, the resultng optmzaton problem s non-convex, the three possble cases descrbed above need to be consdered n detal to maxmze J(p ). Consder p(λ),.e., the sum of the powers transmtted to all users, gven λ. Observng whether p(λ) < PT, or p(λ) > PT provdes us wth the nformaton whether J(p ) s ncreasng or decreasng at p = p (λ): Proposton III.: If p(λ) < PT for a gven λ, J(p) s an ncreasng functon at p = p (λ). Proposton III.: If p(λ) > PT for a gven λ, J(p) s a decreasng functon at p = p (λ). From the prevous two propostons, by observng K p(λ), we can always determne whether J(p) s an ncreasng functon or decreasng functon at p = p (λ). For nstance, f p(λ mn) > P T, J(p ) s a decreasng functon at p = p (λ mn) and the optmum power allocaton s one of three possbltes as descrbed by observaton III.. If p(λ mn) < P T, ths mples p(λ) < PT for λ > λ mn where p (λ) s wthn the concave regon, snce p (λ mn) > p (λ) when U (p ) s ncreasng concave. In ths case, J(p ) s an ncreasng functon when p s wthn the concave regon,.e., 0 p p (λ mn). Hence, no p value wthn the concave regon can be optmum. The optmum power allocaton dctates that p be ether p (λ mn) < p < P T, or P T. Fndng the local optmum soluton wthn the convex regon of U (p ) requres ntensve computatonal complexty. If we fnd a way to dentfy p = P T,.e., the optmalty of TDMA, we can reduce the complexty of dentfyng the optmum polcy. Consder the case of p(λ) < PT for λ where p(λ) s wthn the convex regon of U (p ). In ths case, J(p ) s an ncreasng functon for 0 p < P T. Observaton III.: If p(λ) < PT for λ where p(λ) s wthn the convex regon of U (p ), then p = P T s optmum power allocaton. Observaton III. guarantees the optmalty of TDMA, but stll requres consderable computatonal complexty. Notce that, as descrbed above, when p(λ mn) < P T, the optmum power allocaton does not fall wthn the concave regon of U (p ). By adoptng TDMA as the optmum polcy whenever p(λ mn) < P T, we can sgnfcantly reduce computatonal complexty n fndng local optmum power allocaton n the convex regon of U (p ). In the next secton, we wll term ths short-cut, the near-exact algorthm. The numercal results n Secton IV valdate the accuracy of near-exact algorthm. III.C Algorthms for Transmt Strategy and Power Allocaton Followng the precedng dscusson n Secton III, we conclude that the optmalty of TDMA s determned by checkng p(λ) < PT for all λ where p(λ) s wthn the convex regon of U (p ). Instead of ths potentally computatonally ntensve search, we propose to smply check K p(λ mn) < P T. Clearly, the performance gap between the optmum soluton and the near-optmum soluton utlzng the nearexact algorthm of determnng TDMA optmalty results from the case when the optmum power allocaton s such that p (λ mn) < p < P T. Ths s because the near-exact algorthm decdes that TDMA-mode,.e., p = P T s optmum only f p(λ mn) < P T, regardless of max p J(p ) for p(λ mn) < p P T. In our numercal results, we have observed that such cases,.e., the optmum power allocaton s such that p (λ mn) < p < P T, are rare and that n most cases, the optmum power allocaton s ether 0 < p < p (λ mn) or p = P T. The power allocaton step s done accordng to the mode that s desgnated to be the optmum strategy. When TDMA-mode s optmum, no further power allocaton s necessary. When CDMA-mode s optmum, the optmum power allocaton needs to be found. We should note that observaton III. s a suffcent condton for TDMA optmalty. Hence, even though j= pj(λ mn) > P T, TDMA can be optmum. However, as the orthogonalty factor ncreases, ths condton accounts for almost all of the cases where TDMA s optmum, as demonstrated by numercal results n Secton IV. The followng summarzes the proposed algorthm to maxmze the system utlty, whch we term the system utlty maxmzer, A-SUM.

5 9 9 =0. =0. =0. =0. =0. =0. System Utlty System Utlty (Dscrete) number of users n the system number of users n the system Fg. : System Utlty versus Number of users n the system. Fg. : System Utlty versus Number of users n the system. A-SUM: STEP. If k, then TDMA mode s optmum, STOP. STEP. Fnd λ max for =,, K, and λ mn. STEP. (convex regon) If j= pj(λ mn) < P T, then, TDMA s optmum, STOP. STEP. (concave regon) Fnd the power allocaton n the concave regon J (p (λ)) = max(j(p(λ))), 0 p (λ) p (λ mn) va A-PACR. STEP. Choose max(j (p (λ)), J(P T)). The near-exactness of the algorthm arses from the fact that n STEP, we smply check the suffcent condton for optmalty of TDMA, nstead of solvng for max p J(p ), p (λ mn) < p < P T. The optmum power allocaton (STEP ) entals selectng the users to whch the base staton transmts wth postve power. In other words, f zero power s assgned to one user, that mples that user s not scheduled for transmsson. The algorthm to determne the power allocaton n the concave regon (A-PACR) s summarzed below. It s assumed n A- PACR that j= pj(λ mn) > P T. If ths s not the case, there s no soluton n the concave regon, and A-PACR smply determnes TDMA,.e., p = P T. A-PACR: STEP. Fnd ˆk = arg k max k j= pj(λk max), subject to k j= pj(λk max) P T, k =,, K. STEP. Fnd p j(λ) ( j ˆk) such that ˆk j= pj(λ) = P T where λ mn < λ < λˆk max. The optmum number of users to whch the base staton transmts s selected n STEP. Then, optmum power values are allocated n STEP. Note that a numercal method, e.g., bsecton [] can be used to determne the value of λ. IV. Numercal Results We provde smulaton results for a range of the orthogonalty factor values. Maxmum total power at the base staton s P T = 0 watts. Total nose and ntercell nterference s I = 0 watts. Factor k n the utlty functon s k =.. Thus, when 0., the ndvdual utltes of all users are convex n power, and the TDMA-mode s optmum, regardless of users channel gans. Channel gan from a base staton to moble s modeled as g = r where d d denotes the dstance between moble and the base staton, whch s unformly dstrbuted between 00m and 000m. r s the realzaton of the lognormal fadng coeffcent wth varance db. We have frst examned the accuracy of the near-exact algorthm determnng TDMA optmalty for 0,000 channel gan realzatons. The rato of the near exact algorthm whch correctly determnes the TDMA optmalty n percentage are 9.%, 9.%, 99.9% for = 0., 0., 0., respectvely. As ncreases, as expected, the accuracy ncreases,.e., the case when the optmum power allocaton s such that p (λ mn) p < P T dsappears. For example, when =0., the near-exact algorthm almost always correctly determnes the optmalty of TDMA. Fgure shows the system utlty, gven and the number of users K n the system. The system utlty s averaged over 0,000 channel gan realzatons. The optmum polcy selects the number of users and the power levels for the scheduled users. In general, not all users n the system are smultaneously scheduled. When = 0.,.e., a low orthogonalty factor, smultaneous transmssons do not result n much nterference between users, and the optmum polcy chooses CDMA-mode. However, f the best user has a much hgher channel gan as compared to the rest, the optmum polcy may end up to allocatng all power to that best user. As ncreases, smultaneous transmssons result n hgher nterference, and TDMA-mode becomes the preferred mode so as to avod the nterference. For example, for > 0., TDMAmode s optmum n most cases. For value of > 0., for example, when = 0.,.e., all ndvdual utltes are convex, and TDMA-mode s always optmum. Fgure shows that the

6 system utlty, when = 0., s lower than the system utlty value for smaller values that facltate smultaneous transmssons. The gap between the system utlty for ( < 0.) and the system utlty for = 0. can be nterpreted as the gan of optmum polcy that results n hybrd CDMA/TDMA over the one that results n TDMA, as a result of the dfference n the orthogonalty factor values. Fgure shows the system utlty where the resultng ndvdual utlty s dscretzed to the nteger value. Wth ths, we attempt to nvestgate the performance of the proposed algorthm n a practcal settng where only dscrete rates are avalable. We observe that dscretzaton does not lead to a sgnfcant loss n system utlty especally for large values. Ths s because for large, TDMA results most often, and the quantzaton loss n utlty s due to one user only. Fgure shows the system utlty gan of CDMA-mode over TDMA-mode. When the orthogonalty factor s low, smultaneous transmssons has a gan over the one-at-a-tme transmsson. The gan ncreases as the number of users n the system ncreases up to a certan lmt. As the orthogonalty factor ncreases, however, we observe that the gan dsappears, as eventually TDMA-mode becomes optmum. V. Concluson In ths paper, we consdered the optmum polcy whch conssts of the user(s) the base staton selects to transmt to, and the power allocaton for total utlty maxmzaton n the downlnk. The ndvdual utlty we consdered s the rate to the user. Dependng on the channel condtons,.e., the orthogonalty factor and the users channel gans, the optmum polcy chooses ether TDMA mode or CDMA mode and the correspondng power allocaton. The hgher the orthogonalty factor, the larger the nterference caused by smultaneous transmssons. Ths causes the TDMA-mode beng optmal when the orthogonalty factor s hgh. In contrast, smultaneous transmssons yeld a hgher overall utlty when the orthogonalty factor s low. We observed that dependng on the channel condtons, the overall system utlty may or may not be a concave functon and care must be gven to characterzng the soluton. We took the approach of examnng the system utlty n terms of the best user, and observed propertes that helped us dentfy the optmal polcy. We presented numercal results to support our analytcal fndngs, n partcular to the beneft of allocatng transmsson to smultaneous users n certan scenaros, and evaluate the frequency of TDMA mode optmalty for dfferent system parameters. References [] M. Andrews, K. Kumaran, K. Ramanan, A.L. Stolyar, R. Vjayakumar and P. Whtng. Provdng Qualty of Servce over Shared Wreless Lnk. IEEE Communcatons Magazne, vol. 9, pages 0-, 00. [] P. Bender, P. Black, M. Grob, R. Padovan, N. Sndhushayana and A. Vterb. CDMA/HDR: a bandwdth-effcent hgh-speed wreless data servce for nomadc users. IEEE Communcatons Magazne, pages 0-, 000. [] X. Lu, E. Chong and N. Shroff. Transmsson Schedulng for Effcent Wreless Utlzaton. IEEE Infocom, 00. [] H. Holma. W-CDMA for UMTS:Rado Access for Thrd Generaton Moble Communcatons. Wley and Sons, 00. [] N.B. Mehta, L.J. Greesten, T.M. Wlls and Z. Kostc. Analyss Results for the Orthogonalty Factor n WCDMA Downlnks. IEEE Transactons on Wreless Communcatons, pages - 9, November, 00. System Utlty 0 9 CDMA(=0.) CDMA(=0.) CDMA(=0.) TDMA mode Number of users n the system Fg. : CDMA-mode gan over TDMA-mode for dfferent value. [] C.U. Saraydar, N.B. Mandayam and D.J. Goodman. Prcng and Power Control n a Multcell Wreless Data Network. IEEE Journal on Selected Areas n Communcatons, pages -9, October 00. [] N. Feng, S.C. Mau and N.B. Mandayam. Prcng and Power Control for Jont Network-Centrc and User-Centrc Rado Resource Management. IEEE Transactons on Communcatons, pages -, September 00. [] A. Bedekar, S. Borst, K. Ramanan, P. Whtng and E. Yeh. Downlnk schedulng n CDMA data networks. IEEE Globecom 999. [9] F. Berggren, S.L. Km, R. Juntt and J. Zander. Jont Power Control and Intra-cell Schedulng of DS-CDMA Non-real tme Data. IEEE Journal on Selected Areas n Communcatons, pages 0-0, 00. [0] M.K. Karakayal, R. Yates and L.V. Razoumov. Throughput maxmzaton on the downlnk of a CDMA system. IEEE WCNC 00. [] R. Agrawal and V. Subramanan and R. A. Berry. Jont Schedulng and Resource Allocaton n CDMA Systems. Submtted, 00. Avalable at rberry/opt.pdf. [] X. Qu and K. Chawla. On the Performance of Adaptve Modulaton n Cellular Systems. IEEE Transactons on Communcatons, pages -9, 999. [] S.J. Oh and K.M. Wasserman. Optmalty of Greedy Power Control and Varable Spreadng Gan n Mult-class CDMA Moble Networks. ACM/IEEE Mobcom, 999. [] S. Ulukus and L.J. Greensten. Throughput maxmzaton n CDMA uplnks usng adaptve spreadng and power control. IEEE ISSSTA, 000. [] C. Oh and A. Yener. Downlnk Throughput Maxmzaton for Interference Lmted Multuser Systems: TDMA versus CDMA. submtted to IEEE Transactons on Wreless Communcatons, February 00. [] D.P. Bertsekas. Nonlnear programmng. Athena Scentfc, 99.

2454 IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS, VOL. 6, NO. 7, JULY Changyoon Oh, Member, IEEE, and Aylin Yener, Member, IEEE

2454 IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS, VOL. 6, NO. 7, JULY Changyoon Oh, Member, IEEE, and Aylin Yener, Member, IEEE 25 IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS, VOL., NO. 7, JULY 2007 Downlnk Throughput Maxmzaton for Interference Lmted Multuser Systems: TDMA versus CDMA Changyoon Oh, Member, IEEE, and Ayln Yener,

More information

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

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

Joint Scheduling and Power-Allocation for Interference Management in Wireless Networks

Joint Scheduling and Power-Allocation for Interference Management in Wireless Networks Jont Schedulng and Power-Allocaton for Interference Management n Wreless Networks Xn Lu *, Edwn K. P. Chong, and Ness B. Shroff * * School of Electrcal and Computer Engneerng Purdue Unversty West Lafayette,

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

An 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 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 information

A Lower Bound on SINR Threshold for Call Admission Control in Multiple-Class CDMA Systems with Imperfect Power-Control

A Lower Bound on SINR Threshold for Call Admission Control in Multiple-Class CDMA Systems with Imperfect Power-Control A ower Bound on SIR Threshold for Call Admsson Control n Multple-Class CDMA Systems w Imperfect ower-control Mohamed H. Ahmed Faculty of Engneerng and Appled Scence Memoral Unversty of ewfoundland St.

More information

Simultaneous Optimization of Berth Allocation, Quay Crane Assignment and Quay Crane Scheduling Problems in Container Terminals

Simultaneous Optimization of Berth Allocation, Quay Crane Assignment and Quay Crane Scheduling Problems in Container Terminals Smultaneous Optmzaton of Berth Allocaton, Quay Crane Assgnment and Quay Crane Schedulng Problems n Contaner Termnals Necat Aras, Yavuz Türkoğulları, Z. Caner Taşkın, Kuban Altınel Abstract In ths work,

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

Distributed Non-Autonomous Power Control through Distributed Convex Optimization

Distributed Non-Autonomous Power Control through Distributed Convex Optimization Dstrbuted Non-Autonomous Power Control through Dstrbuted Convex Optmzaton S. Sundhar Ram and V. V. Veeravall ECE Department and Coordnated Scence Lab Unversty of Illnos at Urbana-Champagn Emal: {ssrnv5,vvv}@llnos.edu

More information

Joint Scheduling and Resource Allocation in CDMA Systems

Joint Scheduling and Resource Allocation in CDMA Systems TO APPEAR IEEE TRANSACTIONS ON INFORMATION THEORY Jont Schedulng and Resource Allocaton n CDMA Systems Vjay G. Subramanan, Randall A. Berry, and Rajeev Agrawal Abstract In ths paper, the schedulng and

More information

Joint Scheduling and Resource Allocation in CDMA Systems

Joint Scheduling and Resource Allocation in CDMA Systems TECHNICAL REPORT - JUNE 2009 1 Jont Schedulng and Resource Allocaton n CDMA Systems Vjay G. Subramanan, Randall A. Berry, and Rajeev Agrawal Expanded Techncal Report: A shorter verson of ths paper wll

More information

MAXIMIZING THE THROUGHPUT OF CDMA DATA COMMUNICATIONS THROUGH JOINT ADMISSION AND POWER CONTROL

MAXIMIZING THE THROUGHPUT OF CDMA DATA COMMUNICATIONS THROUGH JOINT ADMISSION AND POWER CONTROL MAXIMIZI THE THROUHPUT OF CDMA DATA COMMUICATIOS THROUH JOIT ADMISSIO AD POWER COTROL Penna Orensten, Davd oodman and Zory Marantz Department of Electrcal and Computer Engneerng Polytechnc Unversty Brooklyn,

More information

Cognitive Access Algorithms For Multiple Access Channels

Cognitive Access Algorithms For Multiple Access Channels 203 IEEE 4th Workshop on Sgnal Processng Advances n Wreless Communcatons SPAWC) Cogntve Access Algorthms For Multple Access Channels Ychuan Hu and Alejandro Rbero, Department of Electrcal and Systems Engneerng,

More information

Joint Scheduling and Resource Allocation in CDMA Systems

Joint Scheduling and Resource Allocation in CDMA Systems ACCEPTED TO IEEE TRANSACTIONS ON INFORMATION THEORY 1 Jont Schedulng and Resource Allocaton n CDMA Systems Vjay G. Subramanan, Randall A. Berry, and Rajeev Agrawal Abstract We consder schedulng and resource

More information

Joint Scheduling and Resource Allocation in CDMA Systems

Joint Scheduling and Resource Allocation in CDMA Systems SUBMITTED TO IEEE TRANSACTIONS ON INFORMATION THEORY 1 Jont Schedulng and Resource Allocaton n CDMA Systems Vjay Subramanan, Randall A. Berry, and Rajeev Agrawal Abstract We consder schedulng and resource

More information

Two-Way and Multiple-Access Energy Harvesting Systems with Energy Cooperation

Two-Way and Multiple-Access Energy Harvesting Systems with Energy Cooperation Two-Way and Multple-Access Energy Harvestng Systems wth Energy Cooperaton Berk Gurakan, Omur Ozel, Jng Yang 2, and Sennur Ulukus Department of Electrcal and Computer Engneerng, Unversty of Maryland, College

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

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

Explicit and Implicit Temperature Constraints in Energy Harvesting Communications

Explicit and Implicit Temperature Constraints in Energy Harvesting Communications Explct and Implct Temperature Constrants n Energy Harvestng Communcatons Abdulrahman Baknna, Omur Ozel 2, and Sennur Ulukus Department of Electrcal and Computer Engneerng, Unversty of Maryland, College

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 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

Assortment Optimization under MNL

Assortment Optimization under MNL Assortment Optmzaton under MNL Haotan Song Aprl 30, 2017 1 Introducton The assortment optmzaton problem ams to fnd the revenue-maxmzng assortment of products to offer when the prces of products are fxed.

More information

Externalities in wireless communication: A public goods solution approach to power allocation. by Shrutivandana Sharma

Externalities in wireless communication: A public goods solution approach to power allocation. by Shrutivandana Sharma Externaltes n wreless communcaton: A publc goods soluton approach to power allocaton by Shrutvandana Sharma SI 786 Tuesday, Feb 2, 2006 Outlne Externaltes: Introducton Plannng wth externaltes Power allocaton:

More information

Optimum Association of Mobile Wireless Devices with a WLAN 3G Access Network

Optimum Association of Mobile Wireless Devices with a WLAN 3G Access Network Optmum Assocaton of Moble Wreless Devces wth a WLAN 3G Access Network K. Premkumar Appled Research Group Satyam Computer Servces Lmted, Bangalore 560 012 Emal: kprem@ece.sc.ernet.n Anurag Kumar Dept. of

More information

Welfare Properties of General Equilibrium. What can be said about optimality properties of resource allocation implied by general equilibrium?

Welfare Properties of General Equilibrium. What can be said about optimality properties of resource allocation implied by general equilibrium? APPLIED WELFARE ECONOMICS AND POLICY ANALYSIS Welfare Propertes of General Equlbrum What can be sad about optmalty propertes of resource allocaton mpled by general equlbrum? Any crteron used to compare

More information

Book Title: Orthogonal Frequency Division Multiple Access. Editors

Book Title: Orthogonal Frequency Division Multiple Access. Editors Book Ttle: Orthogonal Frequency Dvson Multple Access Edtors August 18, 2009 Contents 1 Schedulng and Resource Allocaton n OFDMA 1 1.1 Introducton.................................... 2 1.2 Related Work

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

An Interactive Optimisation Tool for Allocation Problems

An Interactive Optimisation Tool for Allocation Problems An Interactve Optmsaton ool for Allocaton Problems Fredr Bonäs, Joam Westerlund and apo Westerlund Process Desgn Laboratory, Faculty of echnology, Åbo Aadem Unversty, uru 20500, Fnland hs paper presents

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

Chapter 7 Channel Capacity and Coding

Chapter 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 information

EEL 6266 Power System Operation and Control. Chapter 3 Economic Dispatch Using Dynamic Programming

EEL 6266 Power System Operation and Control. Chapter 3 Economic Dispatch Using Dynamic Programming EEL 6266 Power System Operaton and Control Chapter 3 Economc Dspatch Usng Dynamc Programmng Pecewse Lnear Cost Functons Common practce many utltes prefer to represent ther generator cost functons as sngle-

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

Pricing and Resource Allocation Game Theoretic Models

Pricing and Resource Allocation Game Theoretic Models Prcng and Resource Allocaton Game Theoretc Models Zhy Huang Changbn Lu Q Zhang Computer and Informaton Scence December 8, 2009 Z. Huang, C. Lu, and Q. Zhang (CIS) Game Theoretc Models December 8, 2009

More information

Winter 2008 CS567 Stochastic Linear/Integer Programming Guest Lecturer: Xu, Huan

Winter 2008 CS567 Stochastic Linear/Integer Programming Guest Lecturer: Xu, Huan Wnter 2008 CS567 Stochastc Lnear/Integer Programmng Guest Lecturer: Xu, Huan Class 2: More Modelng Examples 1 Capacty Expanson Capacty expanson models optmal choces of the tmng and levels of nvestments

More information

Multiuser Computation Offloading and Downloading for Edge Computing with Virtualization

Multiuser Computation Offloading and Downloading for Edge Computing with Virtualization 1 Multuser Computaton Offloadng and Downloadng for Edge Computng wth Vrtualzaton arxv:1811.07517v1 [cs.it] 19 Nov 2018 Zezu Lang, Yuan Lu, Tat-Mng Lok, and Kabn Huang Abstract Moble-edge computng (MEC)

More information

Econ107 Applied Econometrics Topic 3: Classical Model (Studenmund, Chapter 4)

Econ107 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 information

Perfect Competition and the Nash Bargaining Solution

Perfect Competition and the Nash Bargaining Solution Perfect Competton and the Nash Barganng Soluton Renhard John Department of Economcs Unversty of Bonn Adenauerallee 24-42 53113 Bonn, Germany emal: rohn@un-bonn.de May 2005 Abstract For a lnear exchange

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

Real-Time Systems. Multiprocessor scheduling. Multiprocessor scheduling. Multiprocessor scheduling

Real-Time Systems. Multiprocessor scheduling. Multiprocessor scheduling. Multiprocessor scheduling Real-Tme Systems Multprocessor schedulng Specfcaton Implementaton Verfcaton Multprocessor schedulng -- -- Global schedulng How are tasks assgned to processors? Statc assgnment The processor(s) used for

More information

Hidden Markov Models

Hidden Markov Models Hdden Markov Models Namrata Vaswan, Iowa State Unversty Aprl 24, 204 Hdden Markov Model Defntons and Examples Defntons:. A hdden Markov model (HMM) refers to a set of hdden states X 0, X,..., X t,...,

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

Error Probability for M Signals

Error 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 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

Lecture 4: November 17, Part 1 Single Buffer Management

Lecture 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 information

Difference Equations

Difference 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 information

Economics 101. Lecture 4 - Equilibrium and Efficiency

Economics 101. Lecture 4 - Equilibrium and Efficiency Economcs 0 Lecture 4 - Equlbrum and Effcency Intro As dscussed n the prevous lecture, we wll now move from an envronment where we looed at consumers mang decsons n solaton to analyzng economes full of

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

Lecture 14: Bandits with Budget Constraints

Lecture 14: Bandits with Budget Constraints IEOR 8100-001: Learnng and Optmzaton for Sequental Decson Makng 03/07/16 Lecture 14: andts wth udget Constrants Instructor: Shpra Agrawal Scrbed by: Zhpeng Lu 1 Problem defnton In the regular Mult-armed

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

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

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

More metrics on cartesian products

More 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 information

3.1 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

3.1 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 information

Numerical Heat and Mass Transfer

Numerical Heat and Mass Transfer Master degree n Mechancal Engneerng Numercal Heat and Mass Transfer 06-Fnte-Dfference Method (One-dmensonal, steady state heat conducton) Fausto Arpno f.arpno@uncas.t Introducton Why we use models and

More information

An Improved multiple fractal algorithm

An 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 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

Arizona State University

Arizona State University SCHEDULING AND POWER ALLOCATION TO OPTIMIZE SERVICE AND QUEUE-WAITING TIMES IN COGNITIVE RADIO UPLINKS By arxv:1601.00608v1 [cs.it] 4 Jan 2016 Ahmed Emad Ewasha Commttee: Dr. Chan Tepedelenloğlu, Char

More information

Chapter 13: Multiple Regression

Chapter 13: Multiple Regression Chapter 13: Multple Regresson 13.1 Developng the multple-regresson Model The general model can be descrbed as: It smplfes for two ndependent varables: The sample ft parameter b 0, b 1, and b are used to

More information

A Robust Method for Calculating the Correlation Coefficient

A 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 information

DUE: WEDS FEB 21ST 2018

DUE: 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 information

Polynomial Regression Models

Polynomial Regression Models LINEAR REGRESSION ANALYSIS MODULE XII Lecture - 6 Polynomal Regresson Models Dr. Shalabh Department of Mathematcs and Statstcs Indan Insttute of Technology Kanpur Test of sgnfcance To test the sgnfcance

More information

Module 9. Lecture 6. Duality in Assignment Problems

Module 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 information

Single-Facility Scheduling over Long Time Horizons by Logic-based Benders Decomposition

Single-Facility Scheduling over Long Time Horizons by Logic-based Benders Decomposition Sngle-Faclty Schedulng over Long Tme Horzons by Logc-based Benders Decomposton Elvn Coban and J. N. Hooker Tepper School of Busness, Carnege Mellon Unversty ecoban@andrew.cmu.edu, john@hooker.tepper.cmu.edu

More information

THE 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. 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 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

A PROBABILITY-DRIVEN SEARCH ALGORITHM FOR SOLVING MULTI-OBJECTIVE OPTIMIZATION PROBLEMS

A 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 information

Joint Scheduling and Power Allocations for. Traffic Offloading via Dual-Connectivity

Joint Scheduling and Power Allocations for. Traffic Offloading via Dual-Connectivity Jont Schedulng and Power Allocatons for 1 Traffc Offloadng va Dual-Connectvty Yuan Wu, Yanfe He, Lpng Qan, Janwe Huang, Xuemn (Sherman) Shen arxv:159.9241v1 [cs.ni] 3 Sep 215 Abstract Wth the rapd growth

More information

(1 ) (1 ) 0 (1 ) (1 ) 0

(1 ) (1 ) 0 (1 ) (1 ) 0 Appendx A Appendx A contans proofs for resubmsson "Contractng Informaton Securty n the Presence of Double oral Hazard" Proof of Lemma 1: Assume that, to the contrary, BS efforts are achevable under a blateral

More information

Lecture 12: Discrete Laplacian

Lecture 12: Discrete Laplacian Lecture 12: Dscrete Laplacan Scrbe: Tanye Lu Our goal s to come up wth a dscrete verson of Laplacan operator for trangulated surfaces, so that we can use t n practce to solve related problems We are mostly

More information

Suppose 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

Suppose 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 information

Chapter 7 Channel Capacity and Coding

Chapter 7 Channel Capacity and Coding Chapter 7 Channel Capacty and Codng Contents 7. Channel models and channel capacty 7.. Channel models Bnary symmetrc channel Dscrete memoryless channels Dscrete-nput, contnuous-output channel Waveform

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

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

CHAPTER 5 NUMERICAL EVALUATION OF DYNAMIC RESPONSE

CHAPTER 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 information

Lecture 12: Classification

Lecture 12: Classification Lecture : Classfcaton g Dscrmnant functons g The optmal Bayes classfer g Quadratc classfers g Eucldean and Mahalanobs metrcs g K Nearest Neghbor Classfers Intellgent Sensor Systems Rcardo Guterrez-Osuna

More information

Optimal Resource Allocation in Full-Duplex Wireless-Powered Communication Network

Optimal Resource Allocation in Full-Duplex Wireless-Powered Communication Network 1 Optmal Resource Allocaton n Full-Duplex Wreless-owered Communcaton Network Hyungsk Ju and Ru Zhang, Member, IEEE arxv:143.58v3 [cs.it] 15 Sep 14 Abstract Ths paper studes optmal resource allocaton n

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

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

Markov Chain Monte Carlo Lecture 6

Markov 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 information

princeton univ. F 17 cos 521: Advanced Algorithm Design Lecture 7: LP Duality Lecturer: Matt Weinberg

princeton univ. F 17 cos 521: Advanced Algorithm Design Lecture 7: LP Duality Lecturer: Matt Weinberg prnceton unv. F 17 cos 521: Advanced Algorthm Desgn Lecture 7: LP Dualty Lecturer: Matt Wenberg Scrbe: LP Dualty s an extremely useful tool for analyzng structural propertes of lnear programs. Whle there

More information

DOWNLINK CELL ASSOCIATION OPTIMIZATION FOR HETEROGENEOUS NETWORKS VIA DUAL COORDINATE DESCENT. Kaiming Shen and Wei Yu

DOWNLINK CELL ASSOCIATION OPTIMIZATION FOR HETEROGENEOUS NETWORKS VIA DUAL COORDINATE DESCENT. Kaiming Shen and Wei Yu DOWNLIN CELL ASSOCIATION OPTIMIZATION FOR HETEROGENEOUS NETWORS VIA DUAL COORDINATE DESCENT amng Shen and We Yu Electrcal and Computer Engneerng Dept. Unversty of Toronto, Toronto, Ontaro, Canada M5S 3G4

More information

On balancing multiple video streams with distributed QoS control in mobile communications

On balancing multiple video streams with distributed QoS control in mobile communications On balancng multple vdeo streams wth dstrbuted QoS control n moble communcatons Arjen van der Schaaf, José Angel Lso Arellano, and R. (Inald) L. Lagendjk TU Delft, Mekelweg 4, 68 CD Delft, The Netherlands

More information

Multi-user Multimedia Resource Allocation over. Multi-carrier Wireless Networks

Multi-user Multimedia Resource Allocation over. Multi-carrier Wireless Networks u Mult-user Multmeda Resource Allocaton over Mult-carrer Wreless etworks Y Su * and Mhaela van der Schaar ABSTRACT We address the problem of mult-user vdeo transmsson over the uplnk of mult-carrer networks

More information

Distributed Power Control for Interference-Limited Cooperative Relay Networks

Distributed Power Control for Interference-Limited Cooperative Relay Networks Ths full text paper was peer revewed at the drecton of IEEE Communcatons Socety subject matter experts for publcaton n the IEEE ICC 2009 proceedngs Dstrbuted Power Control for Interference-Lmted Cooperatve

More information

Outline. Communication. Bellman Ford Algorithm. Bellman Ford Example. Bellman Ford Shortest Path [1]

Outline. 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 information

Structure and Drive Paul A. Jensen Copyright July 20, 2003

Structure 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 information

Online Appendix. t=1 (p t w)q t. Then the first order condition shows that

Online Appendix. t=1 (p t w)q t. Then the first order condition shows that Artcle forthcomng to ; manuscrpt no (Please, provde the manuscrpt number!) 1 Onlne Appendx Appendx E: Proofs Proof of Proposton 1 Frst we derve the equlbrum when the manufacturer does not vertcally ntegrate

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

How Bad Is Suboptimal Rate Allocation?

How Bad Is Suboptimal Rate Allocation? How Bad Is Suboptmal Rate Allocaton? Tan Lan, Xaojun Ln 2, Mung Chang, Ruby Lee Department of Electrcal Engneerng, Prnceton Unversty, NJ 08544, USA 2 School of Electrcal and Computer Engneerng, Purdue

More information

Auction-Based Spectrum Sharing

Auction-Based Spectrum Sharing Aucton-Based Spectrum Sharng Janwe Huang, Randall Berry and Mchael L. Hong Department of Electrcal and Computer Engneerng Northwestern Unversty, Evanston IL 60208, USA janwe@northwestern.edu, {rberry,mh}@ece.northwestern.edu

More information

The Multiple Classical Linear Regression Model (CLRM): Specification and Assumptions. 1. Introduction

The 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 information

Portfolios with Trading Constraints and Payout Restrictions

Portfolios with Trading Constraints and Payout Restrictions Portfolos wth Tradng Constrants and Payout Restrctons John R. Brge Northwestern Unversty (ont wor wth Chrs Donohue Xaodong Xu and Gongyun Zhao) 1 General Problem (Very) long-term nvestor (eample: unversty

More information

COS 521: Advanced Algorithms Game Theory and Linear Programming

COS 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 information

On Spatial Capacity of Wireless Ad Hoc Networks with Threshold Based Scheduling

On Spatial Capacity of Wireless Ad Hoc Networks with Threshold Based Scheduling On Spatal Capacty of Wreless Ad Hoc Networks wth Threshold Based Schedulng Yue Lng Che, Ru Zhang, Y Gong, and Lngje Duan Abstract arxv:49.2592v [cs.it] 9 Sep 24 Ths paper studes spatal capacty n a stochastc

More information

Global Sensitivity. Tuesday 20 th February, 2018

Global Sensitivity. Tuesday 20 th February, 2018 Global Senstvty Tuesday 2 th February, 28 ) Local Senstvty Most senstvty analyses [] are based on local estmates of senstvty, typcally by expandng the response n a Taylor seres about some specfc values

More information

A FAST HEURISTIC FOR TASKS ASSIGNMENT IN MANYCORE SYSTEMS WITH VOLTAGE-FREQUENCY ISLANDS

A FAST HEURISTIC FOR TASKS ASSIGNMENT IN MANYCORE SYSTEMS WITH VOLTAGE-FREQUENCY ISLANDS Shervn Haamn A FAST HEURISTIC FOR TASKS ASSIGNMENT IN MANYCORE SYSTEMS WITH VOLTAGE-FREQUENCY ISLANDS INTRODUCTION Increasng computatons n applcatons has led to faster processng. o Use more cores n a chp

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

Analysis of Queuing Delay in Multimedia Gateway Call Routing

Analysis of Queuing Delay in Multimedia Gateway Call Routing Analyss of Queung Delay n Multmeda ateway Call Routng Qwe Huang UTtarcom Inc, 33 Wood Ave. outh Iseln, NJ 08830, U..A Errol Lloyd Computer Informaton cences Department, Unv. of Delaware, Newark, DE 976,

More information

COGNITIVE RADIO NETWORKS BASED ON OPPORTUNISTIC BEAMFORMING WITH QUANTIZED FEEDBACK

COGNITIVE RADIO NETWORKS BASED ON OPPORTUNISTIC BEAMFORMING WITH QUANTIZED FEEDBACK COGNITIVE RADIO NETWORKS BASED ON OPPORTUNISTIC BEAMFORMING WITH QUANTIZED FEEDBACK Ayman MASSAOUDI, Noura SELLAMI 2, Mohamed SIALA MEDIATRON Lab., Sup Com Unversty of Carthage 283 El Ghazala Arana, Tunsa

More information