Energy-Efficient Power Allocation for M2M Communications with Energy Harvesting Transmitter

Size: px
Start display at page:

Download "Energy-Efficient Power Allocation for M2M Communications with Energy Harvesting Transmitter"

Transcription

1 nergy-ffcent ower Allocaton for M2M Communcatons wth nergy Harvestng Transmtter Derrck Wng Kwan Ng and Robert Schober Char of Dgtal Communcatons, The Unverstät rlangen-nürnberg Abstract In ths paper, power allocaton for energy-effcent pont-to-pont machne-to-machne (M2M) communcaton systems wth multple energy harvestng sources s studed. Under a determnstc system settng, we formulate the power allocaton problem as a non-convex optmzaton problem over a fnte horzon takng nto account the crcut energy consumpton, fnte battery storage capactes, and a mnmum requred data rate. The consdered non-convex optmzaton problem s transformed nto a convex optmzaton problem by explotng the propertes of fractonal programmng whch results n an effcent optmal off-lne teratve power allocaton algorthm. In each teraton, the transformed problem s solved by usng dual decomposton and a recursve power allocaton soluton s obtaned for maxmzaton of the energy effcency of data transmsson (bt/joule delvered to the recever). I. INTRODUCTION Recently, a large amount of work has been devoted to machne-to-machne (M2M) communcaton due to ts wde spread applcatons n e-health care, smart cty, and remote montorng, etc. In practce, M2M sensor type devces are usually small and nexpensve whch puts strngent constrants (.e., bandwdth and energy consumpton) on the system desgn [1]. On the other hand, green communcaton has receved much attenton n recent years drven by envronmental concerns [2], [3]. As a result, M2M communcaton systems are not only envsoned to be energy-effcent, but also to be selfsustanable. In the lterature, a tremendous number of green technologes/methods have been proposed for maxmzng the energy effcency (bt-per-joule) of wreless communcaton systems [3]-[7]. Among these technologes, energy harvestng s partcularly appealng and sutable for M2M communcaton snce each M2M devce can harvest energy from natural renewable energy sources such as solar, wnd, and vbraton, etc, thereby reducng substantally the operatng cost of the servce provders. The ntroducton of energy harvestng capabltes nto M2M systems poses many nterestng new challenges for the transmsson desgn due to the tme varyng avalablty of renewable energy sources. In [4] and [5], optmal packet schedulng and power allocaton algorthms were proposed for energy harvestng systems to mnmze the transmsson completon tme, respectvely. However, these works assumed an nfnte battery capacty n the energy harvester and the obtaned results may not be applcable to the case of fnte battery storage. Besdes, they dd not take nto account the crcut energy consumpton and the maxmum energy effcency of these systems s stll unknown even for the case of pontto-pont communcaton. In [6] and [7], the authors proposed an optmal power control tme sequence for maxmzng the throughput by a deadlne wth a sngle energy harvester for dfferent channel scenaros. Yet, the ntermttent nature of energy harvestng of a sngle energy source wll cause the power avalablty at the M2M devce to be hghly random. In other words, such sngle energy harvester desgn may not be able to guarantee the demandng qualty of servce requrements of M2M applcatons such as a mnmum data rate requrement. In ths paper, we address the above ssues. We study the structure of the optmal off lne power allocaton soluton where we assume that non-causal nformaton of channel state nformaton (CSI) and energy arrvals s avalable at the transmtter. The derved off lne soluton consttutes a performance upper bound whch sheds some lght on the desgn of effcent on lne solutons n future research. We formulate the power allocaton problem for energy-effcent M2M communcaton wth multple energy harvesters as an optmzaton problem. By usng nonlnear fractonal programmng, the consdered non-convex optmzaton problem n fractonal form s transformed nto an equvalent optmzaton problem n subtractve form wth a tractable soluton, whch can be found wth an teratve algorthm. In each teraton, dual decomposton s used and a recursve closed-form power allocaton soluton s computed for maxmzaton of the system energy effcency. II. SYSTM MODL We consder a sngle lnk contnuous tme narrowband M2M communcaton system. The transmsson tme s T seconds. We assume that the transmtter adapts the power allocaton L tmes for a gven value of T. Note that the optmal value of L and the tme nstant of each adapton operaton wll be found n the next secton. The data symbol receved at the user at tme nstant t, t T, s gven by y(t) = (t)g(t)h(t)x(t)+z(t), (1) where (t) and x(t) are the transmtted power and the transmtted symbol at tme t, respectvely. h(t) and g(t) are the small scale fadng coeffcent and the path loss between transmtter and recever at tme t, respectvely. z(t) s the addtve whte Gaussan nose (AWGN) at tme t wth zero

2 Transmtter Recever,, 1,1 1,2,,,1,2 Transmtter sgnal processng core nergy harvestng battery1 max1 nergy harvestng battery max ower amplfer Users data nergy flow Jont energy supply Channel gan nergy consumed by the A stemng from energy harvester n (1) = [1] l1,1,2,3,4,5 ε [7] l ε 7 ε [11] l11 poch 1 poch 7 poch 11 n (4) = n (2) =,1 n (6) = n (7) =,2 T,, N,1 N,2 nergy harvestng batteryn maxn Fg. 2. An llustraton of epoches and the meanng of n ( ) for dfferent events and dfferent arrval tme for energy harvester. Fadng changes and energes are harvested at tme nstants denoted by and, respectvely. Fg. 1. A transmtter wth multple energy harvestng sources. mean and varancen W, wheren s the nose power spectral densty and W s the sgnal bandwdth. A. Tme Varyng Fadng Model and nergy Supply Model We adopt a tme varyng system model smlar to the one n [6]. At the transmtter, there are N dfferent types of energy harvesters for supplyng the energy requred for transmsson by the power amplfer (A), cf. Fgure 1. For nstance, the M2M sensor can extract wnd energy or solar energy wth the help of a wnd harvester and a solar energy harvester, respectvely. As a result, the nstantaneous total rado frequency (RF) transmt power at the A n tme nstant t can be wrtten as (t) = (t), t T, (2) where (t) s the nstantaneous power transmtted by the power amplfer, whch s fueled by energy harvester, {1,...,N}. We assume that the changes n the channel gans and energy arrvals n energy harvester are stochastc processes n tme whch can be modeled as osson countng processes wth rates λ F and λ, respectvely [6], [7]. Therefore, changes n the channel gans and the energy arrvals, respectvely, occur n a countable number of tme nstants, whch are ndexed as t F 1,tF 2... and t 1,t 2..., respectvely. The nter-occurrence tmes t F a t F a 1,a {1,2,...}, and t b t b 1,b {1,2,...}, are exponentally dstrbuted wth means 1/λ F and 1/λ, respectvely. Note that we set t = t F = for convenence. A block fadng tme varyng communcaton channel model s consdered. In other words, the fadng level n < t t F 1 s constant but changes to an ndependent value n the next tme nterval t F 1 < t t F 2, and so on. Smlarly,,a unts of energy arrve (to be harvested) at tme t a n energy harvester, cf. Fgure 2. The ncomng energes are collected by the N energy harvesters and buffered n the battery before they are used n data transmsson. On the other hand, we assume that unts of energy arrve/(are avalable) n the battery of energy harvester at t = and the maxmum amount of energy storage n the battery s denoted by max. In the followng, we refer to a change n the channel gans or n the energy level n any one of the batteres as an event and to the tme nterval between two consecutve events as an epoch. Specfcally, epoch a s defned as the tme nterval [t a 1,t a ), where t a 1 and t a are the tmes at whch successve events happen, cf. Fgure 2. B. hyscal Constrants on the nergy Harvesters There are two nherent constrants on each energy harvester: t b δ b 1 C1: ε (u)du,j, b {1,2,...}, (3) d (t) C2: j=,j t j= ε (u)du max, t T,,(4) where,j s the amount of energy harvested by energy harvester at energy arrval j at tme t j. δ s an nfntesmal postve constant for modelng purpose. d (t) = arg max a {t a : t a t} and ε 1 s a constant whch accounts for the neffcency of the A. For nstance, f ε = 5, 5 Watts of power are consumed n the A for every 1 Watts of power radated n the RF and the power effcency s 1 ε = 1 5 = 2%. Constrant C1 mples that n every tme nstant, f the transmtter draws energy from energy harvester to cover the energy requred at the A, t s constraned to use at most the amount of stored energy currently avalable n energy harvester (causalty), although there wll be possbly more energy arrvals n the future. Constrant C2 states that the energy level n energy harvester never exceeds max to prevent the occurrence of an energy overflow n the battery. III. OWR OTIMIZATION ROBLM FORMULATION In the followng, we desgn the power allocaton algorthm based on an nformaton theoretc approach whch nherently assumes that the data buffer at the transmtter s always full 1. 1 In practce, f there s no data n the buffer, the transmtter can smply shut off the A and store all the harvested energy f possble.

3 A. Instantaneous Channel Capacty In ths subsecton, we defne the adopted system performance measure. Gven perfect CSI at the recever, the channel capacty between the transmtter and recever over a transmsson perod of T second(s) wth bandwdth W s gven by T ) C() = W log 2 (1+(t)Γ(t) dt and Γ(t) = g(t) h(t) 2 N W, (5) where Γ(t) s the receved channel gan-to-nose rato (CNR) at the recever at tme t and = { (t),, t T} s the power allocaton polcy. On the other hand, we take nto account the total energy consumpton of the system by ncludng t n the optmzaton objectve functon. For ths purpose, we model the weghted energy dsspaton n the system as the sum of two terms whch can be expressed as U T () = T ε φ (t)dt+ C T, (6) where φ > s a non-negatve constant weght mposed on the use of energy harvester and φ φ k, k. In partcular, the value of φ can be nterpreted as the cost or preference n usng energy harvester. For nstance, f energy harvester explots solar energy, the transmtter may prefer to use battery on sunny days for transmsson by settng φ. On the other hand, C n (6) s the constant requred sgnal processng power 2 at each tme nstant whch ncludes the power dsspatons n the mxer, transmt flter, frequency syntheszer, and dgtal-to-analog converter (DAC), etc. Hence, the weghted energy effcency of the consdered system over a tme perod of T seconds s defned as the total average number of receved bts/joule C() U eff () = U T (). (7) B. Optmzaton roblem Formulaton The optmal power allocaton polcy,, can be obtaned by solvng max U eff () (8) s.t. C1, C2, C3: C() R mn, C4: (t) max, t T, C5: (t),, t T, where C3 specfes the mnmum system data rate requrement R mn. C3 can also be nterpreted as a delay constrant for data transmsson snce at least R mn amount of data has to be transmtted by the end of tme T. In partcle, such constrant 2 We assume that there s a constant energy supply from a non-renewable energy source (e.g. from power grd) for supplyng the energy requred n sgnal processng. Note that we can ncorporate the constant energy supply nto (6) by treatng t as the N-th energy harvester whch has the hghest value of weght φ N. s needed for real tme M2M communcaton servces such as vehcle and asset trackng. Note that although varable R mn n C3 s not an optmzaton varable n ths paper, a balance between energy effcency and system capacty can be struck by varyng R mn. C4 s a constrant on the maxmum transmt power of the transmtter. For nstance, f Zgbee s used for M2M communcaton, the maxmum transmt power s max = 1 W n the US. C5 s the non-negatve constrant on the power allocaton varables. IV. SOLUTION OF TH OTIMIZATION ROBLM The optmzaton problem n (8) s non-convex due to the fractonal form of the objectve functon. We note that there s no standard approach for solvng non-convex optmzaton problems. In order to derve an effcent power allocaton algorthm for the consdered problem, we ntroduce the followng transformaton. A. Transformaton of the Objectve Functon The objectve functon n (8) can be classfed as nonlnear fractonal program [8] and has some nterestng propertes that wll be ntroduced n the followng. Wthout loss of generalty, we defne the maxmum energy effcencyq of the consdered system as q = C( ) U T ( ) = max C() U T (). (9) Then, we can establsh the followng theorem. Theorem 1: The maxmum energy effcency q s acheved f and only f the optmal power allocaton polcy satsfes the followng condton: max C() q U T () (1) = C( ) q U T ( ) =, for C() and U T () >. roof: We can follow a smlar approach as n [9] to prove Theorem 1. The detaled proof s omtted here because of space constrants. By Theorem 1, for any optmzaton problem wth an objectve functon n fractonal form, there exsts an equvalent 3 objectve functon n subtractve form, e.g. C() q U T () n the consdered case. As a result, we can focus on ths equvalent objectve functon n the rest of the paper. B. Iteratve Algorthm for nergy ffcency Maxmzaton In ths secton, we adopt an teratve algorthm (known as the Dnkelbach method) for solvng (8) wth an equvalent objectve functon. The proposed algorthm s summarzed n Table I and the convergence to the optmal energy effcency s guaranteed f we are able to solve the nner problem (11) n each teraton. roof: lease refer to [9] for a proof of convergence. 3 Here, equvalent means that both problem formulatons lead to the same optmal power allocaton polcy.

4 TABL I ITRATIV OWR ALLOCATION ALGORITHM. Algorthm 1 Iteratve ower Allocaton Algorthm 1: Intalze the maxmum number of teratons L max and the maxmum tolerance ɛ 2: Set maxmum energy effcency q = and teraton ndex n = 3: repeat {Man Loop} 4: Solve the nner loop problem n (11) for a gven q and obtan power allocaton polcy { } 5: f C( ) qu T ( ) < ɛ then 6: Convergence = true 7: return { } = { } and q = C( ) U T( ) 8: else 9: Set q = C( ) U T( ) and n = n+1 1: Convergence = false 11: end f 12: untl Convergence = true or n = L max As shown n Table I, n each teraton of the man loop, we solve the followng optmzaton problem for a gven parameter q: max C() qu T () s.t. C1, C2, C3, C4, C5. (11) Soluton of the Man Loop roblem: Although the objectve functon s now n a subtractve form whch s easer to handle, there s stll an obstacle n solvng the above problem. The optmal power allocaton polcy s expected to be tme varyng n the consdered duraton of T seconds. However, t s unclear how often the transmtter should update the power allocaton polcy whch s a hurdle for desgnng a practcal power allocaton algorthm. In order to strke a balance between soluton tractablty and computatonal complexty, we ntroduce the followng lemma whch provdes valuable nsght nto the tme varyng dynamc of the optmal power allocaton polcy. Lemma 1: The optmal power allocaton polcy maxmzng the system energy effcency does not change wthn an epoch. roof: lease refer to the Appendx for a proof of Lemma 1. As revealed by Lemma 1, the optmal power allocaton polcy must be kept constant n each epoch for maxmzng the system energy effcency. As a result, we can dscretze the ntegrals and contnuous varables nvolved n (11). In other words, the number of constrants n (11) reduce to countable quanttes. Wthout loss of generalty, we assume that the channel states change M tmes and energy arrves K tmes n the N energy sources n the duraton of [,T]. Specfcally, we have L = M+K epoch(s) for the consdered duraton of T seconds whch ncludes the epoch caused by at t = for all energy harvesters. Besdes, tme nstant T s treated as an addtonal fadng epoch wth zero channel gan to termnate the process. We defne the length of each epoch as l j = t j t j 1 where epoch j {1,2,...,L} s defned as the tme nterval [t j 1,t j ), cf. Fgure 2. Note that t s defned as t =. For the sake of notatonal smplcty and clarty, we replace the contnuous tme varables wth the correspondng dscrete tme varables,.e., (t) [j], (t) [j], and Γ(t) Γ[j]. Then, the weghted average system throughput and the total weghted energy consumpton can be re-wrtten as C() = l j C[j] and U T () = l j C + l j ε [j]φ, (12) ( respectvely, where C[j] = W log 2 1+( ) N [j])γ[j] s the channel capacty between the transmtter and the recever n epoch l. As a result, the optmzaton problem n (11) s transformed nto the followng convex optmzaton problem: C1: C2: C3: max C() qu T () e e l j ε [j] n [j], e, r r 1 n [j] l j C[j] R mn, εl j [j] max, r, C4: l e [e] l e max, e, C5: [e],,e, (13) where e {1,2,...,L} and r {2,...,L + 1}. In (13), n [j] s defned as the energy whch arrves n epoch j n battery. Hence, n [j] =,a for some a f event j s an energy arrval and n [j] = f event j s a channel gan change, cf. Fgure 2. Now, the transformed problem s jontly concave wth respect to all optmzaton varables 4, and under some mld condtons [1], solvng the dual problem s equvalent to solvng the prmal problem. C. Dual roblem Formulaton In ths subsecton, we solve the power allocaton and schedulng optmzaton problem by solvng ts dual. For ths purpose, we frst need the Lagrangan functon of the prmal problem whch can be wrtten as L(γ,β,ρ,µ,) = l j (1+ρ)C[j] ρr mn ( j γ,j l m ε [m] ( L q l j C + j ) l j ε [j]φ ) n [m] 4 We can follow a smlar approach as n Appendx A to prove the convexty of the above problem for the dscrete tme model.

5 L+1 ( j β,j j=2 j 1 n [m] εl m [m] max ) ( N µ j l j [j] l j max ), (14) where γ s the Lagrange multpler vector assocated wth the causalty constrant C1 n drawng energy from each energy harvester wth elements γ,j, {1,...,N},j {1,...,L}. β s the Lagrange multpler vector correspondng to the maxmum energy level constrant C2 n the battery of the energy harvester wth elements β,j where β,1 =,. ρ s the Lagrange multpler correspondng to the mnmum data rate requrement R mn n C5. µ s the Lagrange multpler vector for constrant C4 on the maxmum power wth elements µ j. Note that the boundary constrants C5 are absorbed nto the Karush-Kuhn-Tucker (KKT) condtons when dervng the optmal soluton n Secton IV-D. Thus, the dual problem s gven by mn max L(γ,β,ρ,µ,). (15) γ,β,ρ,µ D. Dual Decomposton and Sub-roblem Soluton By Lagrange dual decomposton, the dual problem s decomposed nto two parts (nested loops): the frst part (nner loop) s known as sub-problem; the second part (outer loop) s the master problem [9]. Then, the dual problem can be solved teratvely, where n each teraton the transmtter solves the sub-problem (nner loop) by usng KKT condtons for a fxed set of Lagrange multplers, and the master problem (outer loop) s solved usng gradent method. Let [j] denotes the optmal power allocaton soluton of the subproblem for energy harvester n epochj. Wthout loss of generalty, we assume φ 1 < φ 2 <... < φ N for the sake of notatonal smplcty. Usng standard optmzaton technques and the KKT condtons, the optmal power allocatons for the N energy sources n epoch j are gven by the followng recursve equaton: [ ] + 1[j]= W(1+ρ) (ln(2)a 1 [j]) 1 and (16) Γ[j] [ ] + +1 [j]= W(1+ρ) (ln(2)a +1 [j]) 1 Γ[j] d [j],(17) where A [j]= γ,e ε e=j d=1 β,e+1 ε+qφ ε+µ j. (18) e=j The power allocaton solutons n (16) and (17) can be nterpreted as a form of water-fllng. In partcular, varable ρ forces the transmtter to assgn more power for transmsson f the data rate requrement R mn becomes strngent. Interestngly, the optmal values of [j],, have a undrectonal dependence wth each other accordng to the weghts φ,.e., the power drawn from an energy source wth a hgher weght depends on the power drawn from the energy sources wth lesser weghts, but not vce versa. Specfcally, as can be seen n (17),1[j] decreases the water-level n calculatng the value of+1 [j]. In other words, 1 [j] reduces the amount of energy drawn from the less preferable energy sources (hgher values of φ ) for maxmzaton of energy effcency.. Soluton of the Master Dual roblem To solve the master mnmzaton problem n (15),.e., to fnd γ, β, ρ, and µ for a gven, the gradent method can be used snce the dual functon s dfferentable. The gradent update equatons are gven by: [ γ,j (ς +1)= γ,j (ς) ξ 1 (ς) ( j [ β,r (ς +1)= β,r (ς) ξ 2 (ς) ( max n [m] ε [m]l m )] +,,j, (19) r r +,,r,(2) [m]+ εl m [m])] n [ ( L )] +, ρ(ς +1)= ρ(ς) ξ 3 (ς) l j C[j] R mn (21) [ ( µ j (ς +1)= µ j (ς) ξ 4 (ς) max +, j, [j])] (22) where j {1,... L}, r {2,... L}, ndex ς s the teraton ndex, and ξ u (ς), u {1,...,4}, are postve step szes. Then, the updated Lagrange multplers n (19)-(22) are used for solvng the subproblem n (15) va updatng the power allocaton soluton accordng to (16)-(18). Snce the transformed problem n (13) s convex, the dualty gap between dual optmum and prmal optmum s zero and t s guaranteed that the teraton between the master problem and the subproblem converges to the optmal soluton of (11) n the man loop, f the chosen step szes satsfy the nfnte travel condtons [1]. V. RSULTS AND DISCUSSIONS In ths secton, we evaluate the system performance usng smulatons. We assume a transmsson duraton of T = 1 seconds, a carrer center frequency of 2.4 GHz, a sgnal bandwdth of W = 1 khz, a nose power of N W = 134 dbm, and the dstance between transmtter and recever s 5 meters. The small scale fadng coeffcents of the transmtter and recever are generated as Raylegh random varables wth unt varances. The statc crcut power consumpton s set to C = 23 dbm [11], the mnmum data rate requrement of the system s R mn = 2 kbts/s, and the maxmum transmt power s 1 W. The number of energy sources wll be specfed n each case study and each energy harvester has a maxmum energy storage of max = 1 J, and an ntal energy =.5 J n the battery. The amount of energy that can be harvested by each energy harvester n each energy epoch s assumed to be unformly dstrbuted n [,1] J [7]. The channel changes wth rate λ f = 2 ms. On the other hand, we assume a power effcency of 35% n the A,.e., ε = 1.35 =

6 1.2 x Multple energy harvesters dversty gan nergy arrval rates λ roposed algorthm roposed algorthm, 1 teratons, 4 energy harvesters roposed algorthm, 5 teratons, 4 energy harvesters roposed algorthm, 1 teratons, 3 energy harvesters roposed algorthm, 5 teratons, 3 energy harvesters roposed algorthm, 1 teratons, 2 energy harvesters roposed algorthm, 5 teratons, 2 energy harvesters Baselne, 1 teratons Baselne, 5 teratons Average system capacty (kbt/s) roposed algorthm roposed algorthm, 1 teratons, 4 energy harvesters roposed algorthm, 1 teratons, 3 energy harvesters roposed algorthm, 1 teratons, 2 energy harvesters Baselne Baselne 225 Baselne nergy arrval rates λ nergy arrval rates λ Fg. 3. nergy effcency (bt-per-joule) versus energy arrval rate, λ, for the proposed algorthm and the baselne wth dfferent numbers of energy harvesters. Note that f the resource allocator s unable to guarantee the mnmum data rate R mn n T, we set the energy effcency and the average system throughput for that channel realzaton to zero to account for the correspondng falure. The average system energy effcency s obtaned by countng the number of bts whch are successfully decoded by the recever over the total energy consumpton averaged over the mcroscopc fadng. Unless further specfed, n the followng results, the number of teratons refers to the number of teratons of Algorthm 1 n Table I. A. nergy ffcency versus nergy Arrval Rates Fgure 3 llustrates the average energy effcency versus the energy arrval rates, λ, for dfferent numbers of energy harvesters. We defne a vector φ = [φ 1...φ...φ N ]. For the case study of 1, 2, 3, and 4 energy harvester(s), the weght(s) of φ s/are set to φ 1 = [1], φ2 = [.5 1], φ3 = [.1.5 1], and φ 4 = [ ], respectvely, where s a small postve constant for studyng the effect of multple energy harvester dversty. For an energy harvester wth weght φ = 1, a tradtonal contnuous constant energy supply wth an nstantaneous power of 1 W s assumed. The case of φ 1 = [1] s treated as a baselne scheme for comparson. The energy harvesters wth weghts φ < 1, represent some forms of clean energy such as solar energy and wnd energy, etc. The number of teratons for the proposed teratve resource allocaton algorthm s 5 and 1. It can be observed that the performance dfference between 5 and 1 teratons s neglgble whch confrms the practcalty of the proposed algorthm. On the other hand, the growth of energy effcency has a dmnshng return for hgh energy arrval rates. Indeed, when the energy arrval rate ncreases from a small value, the transmtter has a hgher energy level n each battery for performng power allocaton and thus the system energy effcency s enhanced. However, when the arrval rates of energy become exceedngly large, the transmtter s forced to dscharge the batteres n order to prevent a battery overflow, Fg. 4. Average system capacty (kbt/s) versus energy arrval rate, λ, for the proposed algorthm and the baselne. cf. C2 n (4). As a result, the transmtter has to transmt an excess amount of energy for dschargng the batteres whch decreases the system energy effcency gan due to a hgher energy arrval rate. It can be observed that there s an energy effcency gan f we swtch the case from φ 2 to φ 3. Ths s because n φ 3, a more energy effcent source s avalable for transmsson compared to φ 2. Besdes, a form of multple energy harvester dversty can be observed n the energy effcency when we swtch from φ 3 to φ 4. Snce, the performance gan s comng from the transmtter n explotng energy from dfferent energy sources whch changes the ntermttent nature of energy avalablty compared to the case of sngle energy source. On the other hand, the proposed algorthm provdes a sgnfcant performance gan compared to the baselne scheme. Ths s because the baselne scheme can only drawn energy from a less energy-effcent source. B. Average System Capacty versus nergy Arrval Rates Fgure 4 shows the average system capacty versus the energy arrval rates, λ, for dfferent numbers of energy harvesters. We compare the system performance of the proposed algorthm agan wth the baselne scheme. The number of teratons n the proposed algorthm s set to 1. It can be observed that the average system capacty of the proposed algorthm ncreases wth the energy arrval rates. Ths s because more energy s avalable for data transmsson whch results n a capacty gan. We note that, as expected, the baselne scheme acheves a smaller average system capacty than the proposed algorthm snce the proposed algorthm s able to explot energy from dfferent energy sources n T seconds. VI. CONCLUSION In ths paper, we formulated the power allocaton algorthm desgn for a pont-to-pont M2M communcaton systems wth multple energy sources as a non-convex optmzaton

7 problem, n whch the crcut energy consumpton, the fnte battery storage capacty, and the system data rate requrement were taken nto consderaton. By explotng the propertes of nonlnear fractonal programmng, the consdered problem was transformed nto an equvalent convex optmzaton problem wth a tractable soluton. An effcent teratve offlne power allocaton algorthm wth recursve closed-form power allocaton was derved for maxmzaton of the energy effcency. Smulaton results dd not only show that the proposed algorthm converges to the optmal soluton wthn a small number of teratons, but unveled also the achevable maxmum energy effcency. Interestng topcs for future work nclude studyng the optmal on-lne soluton n mult-channel M2M systems. ANDIX - ROOF OF LMMA 1 The proof of Lemma 1 s dvded nto two parts. In the frst part, we prove the convexty of the optmzaton problem n (11). Then, n the second part, we prove a necessary condton for the optmal power allocaton polcy based on the result n part one. 1) roof of the Convexty of the Transformed roblem n (11): We frst consder the concavty of the objectve functon on a per subcarrer bass wth respect to all optmzaton varables. For the sake of notatonal smplcty, we defne the channel capacty between the transmtter and the recever at tme nstant t as C(t) = W log 2 (1+(t)Γ(t)), respectvely. Let the objectve functon n (11) at tme nstanttbef(t,) = C(t) q(ε N φ (t)+ C t). Then, we denote the Hessan matrx of functon f(t,) by H(f(t,)) and the egenvalues of H(f(t,)) by ϕ 1, ϕ 2,..., and ϕ N, respectvely. After some algebrac manpulaton, the egenvalues of H(f(t, )) are gven by ϕ 1 = ϕ 2 =... = ϕ N 1 =, (23) Γ 2 (t)n N ϕ N = (t) ln(2)( N. (24) 2 (t)γ(t)+1) Hence, H(f(t, )) s a negatve sem-defnte matrx snce ϕ. Therefore, f(t,) s jontly concave wth respect to (w.r.t.) optmzaton varables (t) at tme nstant t. Then, the ntegraton of f(t,) over t preserves the concavty of the objectve functon n (11) [1]. On the other hand, the constrants C1-C5 n (11) span a convex feasble set and thus the transformed problem s a concave optmzaton problem. 2) Optmalty of a Constant ower Allocaton olcy n ach poch: Wthout loss of generalty, we consder a tme nterval [t 1,t 2 ) of epoch 1 and a tme nstant τ 1, where t 1 τ 1 < t 2. Suppose an adaptve power allocaton polcy s adopted n t 1 τ 1 < t 2 such that two constant power allocaton polces, { 1 } and { 2 }, are appled n t 1 t < τ 1 and τ 1 t < t 2, respectvely. We assume that { 1 } and { 2 } are feasble solutons to (11) whle 1 2. Now, we defne a thrd power allocaton polcy { 3 } such that 3 = 1(τ 1 t 1)+ 2(t 2 τ 1) t 2 t 1. Note that arthmetc operatons between any two power allocaton polces are defned element-wse. Then, we apply power allocaton polcy 5 { 3 } to the entre epoch 1 and ntegrate f(t,) over tme nterval [t 1,t 2 ) whch yelds: t2 f(t, 3 )dt (a) t2 τ 1 t 1 f(t, 1 )+ t 2 τ 1 f(t, 2 )dt t 1 t 1 t 2 t 1 t 2 t 1 = (τ 1 t 1 )f(t, 1 )+(t 2 τ 1 )f(t, 2 ) = τ1 t 1 f(t, 1 )dt+ t2 τ 1 f(t, 2 )dt, (25) where (a) s due to the concavty of f(t,). In other words, for any adaptve power allocaton polcy wthn an epoch, there always exsts at least one constant power allocaton polcy whch outperforms the adaptve approach. As a result, the optmal power allocaton polcy s non-adaptve wthn each epoch. RFRNCS [1] Y. Zhang, R. Yu, S. Xe, W. Yao, Y. Xao, and M. Guzan, Home M2M Networks: Archtectures, Standards, and QoS Improvement, I Commun. Magazne, vol. 49, pp , Apr [2] Y. Chen, S. Zhang, S. Xu, and G. L, Fundamental Trade-offs on Green Wreless Networks, I Commun. Magazne, vol. 49, pp. 3 37, Jun [3] T. Chen, Y. Yang, H. Zhang, H. Km, and K. Horneman, Network nergy Savng Technologes for Green Wreless Access Networks, I Wreless Commun., vol. 18, pp. 3 38, Oct [4] J. Yang and S. Ulukus, Optmal acket Schedulng n an nergy Harvestng Communcaton System, I Trans. Commun., vol. 6, pp , Jan [5] J. Yang, O. Ozel, and S. Ulukus, Broadcastng wth an nergy Harvestng Rechargeable Transmtter, I Trans. Wreless Commun., vol. 11, pp , Feb [6] O. Ozel, K. Tutuncuoglu, J. Yang, S. Ulukus, and A. Yener, Transmsson wth nergy Harvestng Nodes n Fadng Wreless Channels: Optmal olces, I J. Select. Areas Commun., vol. 29, pp , Sep [7] K. Tutuncuoglu and A. Yener, Optmum Transmsson olces for Battery Lmted nergy Harvestng Nodes, I Trans. Wreless Commun., vol. 11, pp , Mar [8] W. Dnkelbach, On Nonlnear Fractonal rogrammng, Management Scence, vol. 13, pp , Mar [Onlne]. Avalable: [9] D. W. K. Ng,. S. Lo, and R. Schober, nergy-ffcent Resource Allocaton for Secure OFDMA Systems, I Trans. Veh. Technol., preprnt, May 212. [1] S. Boyd and L. Vandenberghe, Convex Optmzaton. Cambrdge Unversty ress, 24. [11] Q. Wang, M. Hempstead, and W. Yang, A Realstc ower Consumpton Model for Wreless Sensor Network Devces, n Thrd Annual I Commun. Socety Conf. on Sensor, Mesh and Ad Hoc Commun. and Networks, vol. 1, Sep. 26, pp ower allocaton polcy { 3 } s also a feasble soluton to (11) by the convexty of the feasble soluton set.

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

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

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

More information

Lecture 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

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

Solutions to exam in SF1811 Optimization, Jan 14, 2015

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

More information

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

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

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

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

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

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

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

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

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

More information

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

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

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

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

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

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

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

An Admission Control Algorithm in Cloud Computing Systems

An Admission Control Algorithm in Cloud Computing Systems An Admsson Control Algorthm n Cloud Computng Systems Authors: Frank Yeong-Sung Ln Department of Informaton Management Natonal Tawan Unversty Tape, Tawan, R.O.C. ysln@m.ntu.edu.tw Yngje Lan Management Scence

More information

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

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

More information

VQ widely used in coding speech, image, and video

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

More information

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

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

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

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

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

4884 IEEE TRANSACTIONS ON COMMUNICATIONS, VOL. 61, NO. 12, DECEMBER Energy Cooperation in Energy Harvesting Communications

4884 IEEE TRANSACTIONS ON COMMUNICATIONS, VOL. 61, NO. 12, DECEMBER Energy Cooperation in Energy Harvesting Communications 4884 IEEE TRANSACTIONS ON COMMUNICATIONS, VOL. 6, NO., DECEMBER 3 Energy Cooperaton n Energy Harvestng Communcatons Ber Guraan, Student Member, IEEE, Omur Ozel, Student Member, IEEE, Jng Yang, Member,

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

Natural Language Processing and Information Retrieval

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

More information

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

Digital Modems. Lecture 2

Digital Modems. Lecture 2 Dgtal Modems Lecture Revew We have shown that both Bayes and eyman/pearson crtera are based on the Lkelhood Rato Test (LRT) Λ ( r ) < > η Λ r s called observaton transformaton or suffcent statstc The crtera

More information

IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS, VOL. 33, NO. 3, MARCH

IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS, VOL. 33, NO. 3, MARCH IEEE JOURAL O SELECTED AREAS I COMMUICATIOS, VOL. 33, O. 3, MARCH 205 467 Optmum Polces for an Energy Harvestng Transmtter Under Energy Storage Losses Kaya Tutuncuoglu, Student Member, IEEE, Ayln Yener,

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

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

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

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

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

More information

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

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

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

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

An Integrated OR/CP Method for Planning and Scheduling

An Integrated OR/CP Method for Planning and Scheduling An Integrated OR/CP Method for Plannng and Schedulng John Hooer Carnege Mellon Unversty IT Unversty of Copenhagen June 2005 The Problem Allocate tass to facltes. Schedule tass assgned to each faclty. Subect

More information

Chapter - 2. Distribution System Power Flow Analysis

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

More information

Amiri s Supply Chain Model. System Engineering b Department of Mathematics and Statistics c Odette School of Business

Amiri s Supply Chain Model. System Engineering b Department of Mathematics and Statistics c Odette School of Business Amr s Supply Chan Model by S. Ashtab a,, R.J. Caron b E. Selvarajah c a Department of Industral Manufacturng System Engneerng b Department of Mathematcs Statstcs c Odette School of Busness Unversty 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

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

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

More information

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

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

More information

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

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

More information

The Geometry of Logit and Probit

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

More information

Energy Efficient Resource Allocation in Machine-to-Machine Communications with Multiple Access and Energy Harvesting for IoT

Energy Efficient Resource Allocation in Machine-to-Machine Communications with Multiple Access and Energy Harvesting for IoT 1 Energy Effcent Resource Allocaton n Machne-to-Machne Communcatons wth Multple Access and Energy Harvestng for IoT Zhaohu Yang, We Xu, Senor Member, IEEE, Yjn Pan, Cunhua Pan, and Mng Chen arxv:1711.10776v1

More information

A Simple Inventory System

A Simple Inventory System A Smple Inventory System Lawrence M. Leems and Stephen K. Park, Dscrete-Event Smulaton: A Frst Course, Prentce Hall, 2006 Hu Chen Computer Scence Vrgna State Unversty Petersburg, Vrgna February 8, 2017

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 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

Tokyo Institute of Technology Periodic Sequencing Control over Multi Communication Channels with Packet Losses

Tokyo Institute of Technology Periodic Sequencing Control over Multi Communication Channels with Packet Losses oyo Insttute of echnology Fujta Laboratory oyo Insttute of echnology erodc Sequencng Control over Mult Communcaton Channels wth acet Losses FL6-7- /8/6 zwrman Gusrald oyo Insttute of echnology Fujta Laboratory

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

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

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

More information

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

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

Convexity preserving interpolation by splines of arbitrary degree

Convexity preserving interpolation by splines of arbitrary degree Computer Scence Journal of Moldova, vol.18, no.1(52), 2010 Convexty preservng nterpolaton by splnes of arbtrary degree Igor Verlan Abstract In the present paper an algorthm of C 2 nterpolaton of dscrete

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

Energy Efficient Resource Allocation for Quantity of Information Delivery in Parallel Channels

Energy Efficient Resource Allocation for Quantity of Information Delivery in Parallel Channels TRANSACTIONS ON EMERGING TELECOMMUNICATIONS TECHNOLOGIES Trans. Emergng Tel. Tech. 0000; 00: 6 RESEARCH ARTICLE Energy Effcent Resource Allocaton for Quantty of Informaton Delvery n Parallel Channels Jean-Yves

More information

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

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

More information

A SEPARABLE APPROXIMATION DYNAMIC PROGRAMMING ALGORITHM FOR ECONOMIC DISPATCH WITH TRANSMISSION LOSSES. Pierre HANSEN, Nenad MLADENOVI]

A SEPARABLE APPROXIMATION DYNAMIC PROGRAMMING ALGORITHM FOR ECONOMIC DISPATCH WITH TRANSMISSION LOSSES. Pierre HANSEN, Nenad MLADENOVI] Yugoslav Journal of Operatons Research (00) umber 57-66 A SEPARABLE APPROXIMATIO DYAMIC PROGRAMMIG ALGORITHM FOR ECOOMIC DISPATCH WITH TRASMISSIO LOSSES Perre HASE enad MLADEOVI] GERAD and Ecole des Hautes

More information

Clock Synchronization in WSN: from Traditional Estimation Theory to Distributed Signal Processing

Clock Synchronization in WSN: from Traditional Estimation Theory to Distributed Signal Processing Clock Synchronzaton n WS: from Tradtonal Estmaton Theory to Dstrbuted Sgnal Processng Yk-Chung WU The Unversty of Hong Kong Emal: ycwu@eee.hku.hk, Webpage: www.eee.hku.hk/~ycwu Applcatons requre clock

More information

Minimisation of the Average Response Time in a Cluster of Servers

Minimisation of the Average Response Time in a Cluster of Servers Mnmsaton of the Average Response Tme n a Cluster of Servers Valery Naumov Abstract: In ths paper, we consder task assgnment problem n a cluster of servers. We show that optmal statc task assgnment s tantamount

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

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

6.854J / J Advanced Algorithms Fall 2008

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

More information

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

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

Queueing Networks II Network Performance

Queueing Networks II Network Performance Queueng Networks II Network Performance Davd Tpper Assocate Professor Graduate Telecommuncatons and Networkng Program Unversty of Pttsburgh Sldes 6 Networks of Queues Many communcaton systems must be modeled

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

Joint Energy Management and Resource Allocation in Rechargable Sensor Networks

Joint Energy Management and Resource Allocation in Rechargable Sensor Networks Jont Energy Management and Resource Allocaton n Rechargable Sensor Networks Ren-Shou Lu, Prasun Snha and C. Emre Koksal Department of CSE and ECE The Oho State Unversty Envronmental Energy Harvestng Many

More information

State Amplification and State Masking for the Binary Energy Harvesting Channel

State Amplification and State Masking for the Binary Energy Harvesting Channel State Amplfcaton and State Maskng for the Bnary Energy Harvestng Channel Kaya Tutuncuoglu, Omur Ozel 2, Ayln Yener, and Sennur Ulukus 2 Department of Electrcal Engneerng, The Pennsylvana State Unversty,

More information

Equilibrium with Complete Markets. Instructor: Dmytro Hryshko

Equilibrium with Complete Markets. Instructor: Dmytro Hryshko Equlbrum wth Complete Markets Instructor: Dmytro Hryshko 1 / 33 Readngs Ljungqvst and Sargent. Recursve Macroeconomc Theory. MIT Press. Chapter 8. 2 / 33 Equlbrum n pure exchange, nfnte horzon economes,

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

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

Lecture 20: November 7

Lecture 20: November 7 0-725/36-725: Convex Optmzaton Fall 205 Lecturer: Ryan Tbshran Lecture 20: November 7 Scrbes: Varsha Chnnaobreddy, Joon Sk Km, Lngyao Zhang Note: LaTeX template courtesy of UC Berkeley EECS dept. Dsclamer:

More information

Statistical Circuit Optimization Considering Device and Interconnect Process Variations

Statistical Circuit Optimization Considering Device and Interconnect Process Variations Statstcal Crcut Optmzaton Consderng Devce and Interconnect Process Varatons I-Jye Ln, Tsu-Yee Lng, and Yao-Wen Chang The Electronc Desgn Automaton Laboratory Department of Electrcal Engneerng Natonal Tawan

More information

CHAPTER 7 STOCHASTIC ECONOMIC EMISSION DISPATCH-MODELED USING WEIGHTING METHOD

CHAPTER 7 STOCHASTIC ECONOMIC EMISSION DISPATCH-MODELED USING WEIGHTING METHOD 90 CHAPTER 7 STOCHASTIC ECOOMIC EMISSIO DISPATCH-MODELED USIG WEIGHTIG METHOD 7.1 ITRODUCTIO early 70% of electrc power produced n the world s by means of thermal plants. Thermal power statons are the

More information

FUZZY GOAL PROGRAMMING VS ORDINARY FUZZY PROGRAMMING APPROACH FOR MULTI OBJECTIVE PROGRAMMING PROBLEM

FUZZY GOAL PROGRAMMING VS ORDINARY FUZZY PROGRAMMING APPROACH FOR MULTI OBJECTIVE PROGRAMMING PROBLEM Internatonal Conference on Ceramcs, Bkaner, Inda Internatonal Journal of Modern Physcs: Conference Seres Vol. 22 (2013) 757 761 World Scentfc Publshng Company DOI: 10.1142/S2010194513010982 FUZZY GOAL

More information

Singular Value Decomposition: Theory and Applications

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

More information

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

Notes prepared by Prof Mrs) M.J. Gholba Class M.Sc Part(I) Information Technology

Notes prepared by Prof Mrs) M.J. Gholba Class M.Sc Part(I) Information Technology Inverse transformatons Generaton of random observatons from gven dstrbutons Assume that random numbers,,, are readly avalable, where each tself s a random varable whch s unformly dstrbuted over the range(,).

More information

1 Convex Optimization

1 Convex Optimization Convex Optmzaton We wll consder convex optmzaton problems. Namely, mnmzaton problems where the objectve s convex (we assume no constrants for now). Such problems often arse n machne learnng. For example,

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

MIMO Systems and Channel Capacity

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

More information

Joint Scheduling of Rate-guaranteed and Best-effort Services over a Wireless Channel

Joint Scheduling of Rate-guaranteed and Best-effort Services over a Wireless Channel Jont Schedulng of Rate-guaranteed and Best-effort Servces over a Wreless Channel Murtaza Zafer and Eytan Modano Abstract We consder mult-user schedulng over the downln channel n wreless data systems. Specfcally,

More information

Annexes. EC.1. Cycle-base move illustration. EC.2. Problem Instances

Annexes. 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 information

10) Activity analysis

10) Activity analysis 3C3 Mathematcal Methods for Economsts (6 cr) 1) Actvty analyss Abolfazl Keshvar Ph.D. Aalto Unversty School of Busness Sldes orgnally by: Tmo Kuosmanen Updated by: Abolfazl Keshvar 1 Outlne Hstorcal development

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

Generalized Linear Methods

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

More information

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

e - c o m p a n i o n

e - c o m p a n i o n OPERATIONS RESEARCH http://dxdoorg/0287/opre007ec e - c o m p a n o n ONLY AVAILABLE IN ELECTRONIC FORM 202 INFORMS Electronc Companon Generalzed Quantty Competton for Multple Products and Loss of Effcency

More information

Appendix for Causal Interaction in Factorial Experiments: Application to Conjoint Analysis

Appendix for Causal Interaction in Factorial Experiments: Application to Conjoint Analysis A Appendx for Causal Interacton n Factoral Experments: Applcaton to Conjont Analyss Mathematcal Appendx: Proofs of Theorems A. Lemmas Below, we descrbe all the lemmas, whch are used to prove the man theorems

More information

NUMERICAL DIFFERENTIATION

NUMERICAL 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 information

Which Separator? Spring 1

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

More information

Department of Electrical & Electronic Engineeing Imperial College London. E4.20 Digital IC Design. Median Filter Project Specification

Department of Electrical & Electronic Engineeing Imperial College London. E4.20 Digital IC Design. Median Filter Project Specification Desgn Project Specfcaton Medan Flter Department of Electrcal & Electronc Engneeng Imperal College London E4.20 Dgtal IC Desgn Medan Flter Project Specfcaton A medan flter s used to remove nose from a sampled

More information

Errors for Linear Systems

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

More information