Closed form solutions for water-filling problems in optimization and game frameworks

Size: px
Start display at page:

Download "Closed form solutions for water-filling problems in optimization and game frameworks"

Transcription

1 Closed form solutons for water-fllng problems n optmzaton and game frameworks Etan Altman INRIA BP93 24 route des Lucoles 692 Sopha Antpols FRANCE altman@sopha.nra.fr Konstantn Avrachenkov INRIA BP93 24 route des Lucoles 692 Sopha Antpols FRANCE k.avrachenkov@sopha.nra.fr Andrey Garnaev St. Petersburg State Unversty Unverstetsk pr 35 Peterhof St Petersburg 954 RUSSIA agarnaev@rambler.ru ABSTRACT We study power control n optmzaton and game frameworks. In the optmzaton framework there s a sngle decson maker who assgns network resources and n the game framework players share the network resources accordng to Nash equlbrum. The soluton of these problems s based on so-called water-fllng technque, whch n turn uses bsecton method for soluton of non-lnear equatons for Lagrange multples. Here we provde a closed form soluton to the water-fllng problem, whch allows us to solve t n a fnte number of operatons. Also, we produce a closed form soluton for the Nash equlbrum n symmetrc Gaussan nterference game. In addton, to ts mathematcal beauty, the explct soluton allows one to study lmtng cases when the crosstalk coeffcent s ether small or large. We provde an alternatve smple proof of the convergence of the Iteratve Water Fllng Algorthm. Furthermore, t turns out that the convergence of Iteratve Water Fllng Algorthm slows down when the crosstalk coeffcent s large. Usng the closed form soluton, we can avod ths problem. Fnally, we compare the non-cooperatve approach wth the cooperatve approach and show that the non-cooperatve approach results n a more far resource dstrbuton.. INTRODUCTION In wreless networks and DSL access networks the total avalable power for sgnal transmsson has to be dstrbuted among several resources. In the context of wreless networks, the resources may correspond to frequency bands (e.g. as n OFDM), or they may correspond to capacty avalable at dfferent tme slots. In the context of DSL access networks, the resources correspond to avalable frequency tones. Ths spectrum of problems can be consdered n ether optmza- Ths work was partly supported by BoNets European project and by the jont RFBR and NNSF Grant no Permssontomakedgtalorhardcopesofallorpartofthsworkfor personal or classroom use s granted wthout fee provded that copes are not made or dstrbuted for proft or commercal advantage and that copes bearthsnotceandthefullctatononthefrstpage.tocopyotherwse,to republsh, to post on servers or to redstrbute to lsts, requres pror specfc permsson and/or a fee. GameComm 7, October 22, 27, Nantes, France Copyrght 27 ICST ton scenaro or as a result of a non-cooperatve game scenaro. The optmzaton scenaro leads to Water Fllng Optmzaton Problem [4, 6, 3] and the game scenaro leads to Water Fllng Game or Gaussan Interference Game [,, 2, 4]. In the optmzaton scenaro, one needs to maxmze a concave functon (Shannon capacty) subject to power constrants. The Lagrange multpler correspondng to the power constrant s determned by a non-lnear equaton. In the prevous works [4, 6, 3], t was suggested to fnd the Lagrange multpler by means of a bsecton algorthm, comes the name Water Fllng Problem. Here we show that the Lagrange multpler and hence the optmal soluton of the water fllng problem can be found n explct form wth a fnte number of operatons. In the multuser context, one can vew the problem n ether cooperatve or non-cooperatve settng. If a centralzed controller wants to maxmze the sum of all users rates, the controller wll face a non-convex optmzaton problem [3]. On the other hand, n the non-cooperatve settng, the power allocaton problem becomes a game problem each user perceves the sgnals of the other users as nterference and maxmzes a concave functon of the nose to nterference rato. In [, 4] the spectrum of avalable resources was contnuous, here as n [,, 2] we consder the dscrete spectrum of avalable resources. A natural approach n the non-cooperatve settng s the applcaton of the Iteratve Water Fllng Algorthm (IWFA) [5]. Recently, the authors of [] proved the convergence of IWFA under farly general condtons. In the present work we consder the case of symmetrc water fllng game wth two users. The restrcton to the symmetrc scenaro allows us to fnd Nash equlbrum n explct form. In addton, to ts mathematcal beauty, the explct soluton allows one to fnd the Nash equlbrum n water fllng game n a fnte number of operatons and to study lmtng cases when the crosstalk coeffcent s ether small or large. As a by-product, we obtan an alternatve smple proof of the convergence of the Iteratve Water Fllng Algorthm. Furthermore, t turns out that the convergence of IWFA slows down when the crosstalk coeffcent s large. Usng the closed form soluton, we can avod ths problem. Fnally, we compare the non-cooperatve approach wth the cooperatve approach and conclude that the cost of anarchy s small n the case of small crosstalk coeffcents. We also show that the non-cooperatve approach results n a more far resource dstrbuton. Applcatons that can mostly beneft from decentralzed non-cooperatve power control are ad-hoc and sensor

2 networks wth no predefned base statons [5, 9, 7]. An nterested reader can fnd more references on non-cooperatve power control n [2, ]. We would lke to menton that the water fllng problem and jammng games wth transmsson costs have been analyzed n []. The structure of the paper s as follows: In the next Secton 2, we recall the sngle decson maker setup of the water fllng optmzaton problem. Then, n Secton 3 we provde ts explct soluton. In Secton 4 we formulate two user symmetrc water fllng game and characterze ts Nash equlbrum. In Secton 5 we gve an alternatve smple proof of the convergence of the teratve water fllng algorthm. In Secton 6 we gve the explct form of the players strategy n the Nash equlbrum. In Secton 7 we confrm our fndng wth the help of numercal example. In that secton we also show that the cost of anarchy s small when the crosstalk coeffcent s small. We make conclusons n Secton. 2. SINGLE DECISION MAKER We consder the followng power allocaton problem n the case of a sngle decson maker. There s a sngle decson maker (also called Transmtter ) who wants to send nformaton usng n ndependent resources so as to maxmze the Shannon capacty. We further assume that resource has a weght of π. Possble nterpretatons: () The resources may correspond to capacty avalable at dfferent tme slots; we assume that there s a varyng envronment whose state changes among a fnte set of states [, n], accordng to some ergodc stochastc process wth statonary dstrbuton {π } n =. We assume that both players have perfect knowledge of the envronment state at the begnnng of each tme slot. () The resources may correspond to frequency bands (e.g. as n OFDM) one should assgn dfferent power levels for dfferent sub-carrers [3]. In that case we may take π = /n for all. P The strategy of Transmtter s T = (T,..., T n) wth n = πt = T, T, π > for [, n] and T >. As the payoff to Transmtter we take the Shannon capacty v(t) = π ln = + T N, N > s the nose level n the sub-carrer. Defne also the functon H T(ω) = π T (ω). = We would lke to emphasze that the above generalzed descrpton of the water-fllng problem can be used for power allocaton n tme as well as power allocaton n space-frequency. Followng the standard water-fllng approach [4, 6, 3], whch assumes applcaton Kuhn-Tacker Theorem, we have the followng result. Theorem. Let T (ω) = ˆ/ω N for [, n]. + Then T(ω ) = (T (ω ),..., T n(ω )) s the unque optmal strategy and ts payoff s v(t(ω )) ω s the unque root of the equaton H T(ω) = T. () We observe that the above result can be easly proved wthout applcaton of the Kuhn-Tacker condtons. In the appendx we supply ths proof. 3. CLOSED FORM SOLUTION FOR WATER FILLING PROBLEM In the prevous studes of the water-fllng problems t was suggested to use numercal (e.g., bsecton) method to solve equaton (). Here we propose an explct form approach for the soluton of equaton (). Wthout loss of generalty we can assume that the subcarrers are arranged by the nose level as follows: N N Nn Theorem 2. The soluton of the water-fllng optmzaton problem s gven by k T + π t(nt N ) >< T t=, f k, = k (2) π t t=, f > k, k can be found from the followng condtons: ϕ t = ϕ k < T ϕ k+, (3) t π (Nt N ) for t [, n]. = Let us demonstrate the closed form approach by a numercal example. Take n = 5, T =, N = κ, κ =.7, π = /5 for [, 5]. Then, as the frst step we calculate ϕ t for t [, 5]. In our case we get (..4,.66,.29, 4.5). Then, by (3), k = 3. Thus, by (2), the optmal water-fllng strategy s T = (2.53,.3,.64,, ) wth payoff SYMMETRIC WATER FILLING GAME Two Transmtters try to send nformaton through n resources so as to maxmze the qualty of the transmtted nformaton. The strategy of Transmtter j s T j = (T j,..., T j n) wth T j and wth n = π T j = T j, (4) T j > for j =, 2. The payoffs to Transmtters are gven as follows v (T, T 2 T ) = π ln + gt 2 = +, N v 2 (T, T 2 T 2 ) = π ln + gt +, N = g (, ). These payoffs correspond to Shannon capactes. Ths s an nstance of the Water-Fllng or Gaussan Interference Game [, 4, 5]. In the mportant partcular cases of OFDM wreless network and DSL access network,

3 π = /n, =,..., n. In ths work we restrct ourselves to the case of symmetrc game wth equal crosstalk coeffcents. Ths stuaton can correspond for example to the scenaro when the transmtters are stuated at about the same dstance from the base staton. We shall characterze a Nash Equlbrum of ths problem. The strateges (T, T 2 ) consttute a Nash Equlbrum, f for any strateges (T, T 2 ) the followng nequaltes hold: v (T, T 2 ) v (T, T 2 ), v 2 (T, T 2 ) v 2 (T, T 2 ). Snce v and v 2 are concave n T and T 2 respectvely, the Kuhn-Tucker condtons mply the followng theorem. Theorem 3. (T, T 2 ) s a Nash equlbrum f and only f there are non-negatve ω and ω 2 (Lagrange multplers) such that ( = ω j for T j >, gt m + T j + N j, m {, 2} and m j. ω j for T j =, We would lke to emphasze that the postvty of the Lagrange multplers ω and ω 2 s specfc for our problem. In general, the Lagrange multplers correspondng to equalty constrants do not have any restrctons on the sgn. Let us ntroduce the followng sets: I (ω, ω 2 ) = { [, n] : ω N, ω 2 N }, I (ω, ω 2 ) = { [, n] : N < ω and ether N ω 2 or N < ω 2 and ω 2 N g( ω N )}, I (ω, ω 2 ) = { [, n] : N < ω 2 and ether N ω or N < ω and ω N g( ω 2 N )}, I (ω, ω 2 ) = { [, n] : < g( ω 2 N ) < ω N < g ( ω 2 N )}. The next result characterzes the forms that the Nash equlbrum can take. Lemma. Let (T, T 2 ) be a Nash equlbrum, then () f T () f T () f T (v) f T = and T 2 = then I (ω, ω 2 ), > and T 2 = and T 2 > and T 2 = then I (ω, ω 2 ) and T = /ω N, > then I (ω, ω 2 ) and T 2 = /ω 2 N, > then I (ω, ω 2 ) and T = (/ω N ) g(/ω 2 N ) g 2, T 2 = (/ω2 N ) g(/ω N ) g 2. (5) (6) Although the game has symmetrc nature there are some non-symmetrc features mpacted by the fact that the Lagrange multples are dfferent as a rule. Ths dfference wll allow us to smplfy the structure of the sets I and the strateges. For ths purpose frst ntroduce the followng auxlary notatons for postve ω and ω 2 : () f ω < ω 2, so /ω 2 < /ω then let I (ω, ω 2 ) = { [, n] : ω N }, I (ω, ω 2 /ω 2 g/ω ) = { [, n] : N < g ω }, I (ω, ω 2 ) =, I (ω, ω 2 ) = { [, n] : N < /ω2 g/ω }, g () f ω 2 < ω, so /ω < /ω 2 then let I (ω, ω 2 ) = { [, n] : I (ω, ω 2 ) =, I (ω, ω 2 ) = { [, n] : ω 2 N }, /ω g/ω 2 g N < ω 2 }, I (ω, ω 2 ) = { [, n] : N < /ω g/ω 2 }, g () f ω 2 = ω then let I (ω, ω 2 ) = { [, n] : ω 2 N }, I (ω, ω 2 ) =, I (ω, ω 2 ) =, I (ω, ω 2 ) = { [, n] : N < ω 2 }. The next lemma asserts that the sets I are concdes wth the sets I. Lemma 2. There are the followng relatons between sets I and I: I (ω, ω 2 ) = I (ω, ω 2 ), I (ω, ω 2 ) = I (ω, ω 2 ), I (ω, ω 2 ) = I (ω, ω 2 ) and I (ω, ω 2 ) = I (ω, ω 2 ) for postve ω and ω 2. Now we ntroduce some strateges, whch the Nash Equlbrum wll have the form of. Namely, for postve ω and ω 2 such that ω ω 2 and for [, n] we ntroduce the followng notatons: T (ω, ω 2 ) = B + ω g ω 2 g N C A f N < >< ω N f f f ω 2 g ω g ω N, ω 2 g ω g, N < ω, T 2 (ω, ω 2 ) = B >< + ω 2 g ω g N C A f N < /ω2 g/ω g, ω 2 g ω g N,

4 ether n the followng equvalent form as follows T 2 (ω, ω 2 ) ( `t2 = + g N f N < t 2, f t 2 N, T (ω, ω 2 ) >< + g `( + g)t gt 2 N f N < t 2, = t N f t 2 N < t, f t N, It s clear that t 2 = /ω2 g/ω, t = g ω. (7) /ω = t 2, /ω 2 = gt + ( g)t 2. () For the case ω > ω 2, T (ω, ω 2 ) and T 2 (ω, ω 2 ) can be defned by symmetry. The next result smplfes the form of the Nash equlbrum, gven by Lemma and t shows that the strateges are not so symmetrc as t could be expected and ther nonsymmetrc structure motvated by dfference n Lagrange multplers and so n the power of the sgnals the players have to transfer. Theorem 4. Each Nash equlbrum s of the form (T (ω, ω 2 ), T 2 (ω, ω 2 )) for some postve ω and ω 2. The next result shows that there s a monotonous dependence between the power of the sgnals the players have to transfer and Lagrange multplers. Corollary. Let (T (ω, ω 2 ), T 2 (ω, ω 2 )) be a Nash equlbrum. If T > T 2, then ω < ω 2. To fnd the equlbrum strateges we have to fnd ω and ω 2 such that the followng condtons hold H j (ω, ω 2 ) = H j (ω, ω 2 ) = T j for j =, 2 (9) π T j (ω, ω 2 ) for j =, 2. = The next Lemma shows that ths system has the unque soluton and moreover ts proof supples a smple method of ts determnaton. Lemma 3. The system of non-lnear equatons (9) has unque postve soluton (ω, ω 2 ). Proof. Wthout lost of generalty we can assume that T T 2. Let (ω, ω 2 ) be the postve soluton of (9). Then, by Corollary, ω < ω 2. Thus, nstead of the system of equaton (9) wth varables ω and ω 2 we can consder the followng equvalent system of equaton () wth varables t and t 2 < t 2 t : H 2 (t 2 ) = T 2, H (t, t 2 ) = T, () H (t, t 2 ) = H 2 (t 2 ) = + g {:t 2 N <t } + {:N <t2 } {:t 2 >N } π (t 2 N ), π (t N ) π + g `( + g)t gt 2 N. It s clear that H 2 ( ) s contnuous n (, ), H2 (τ) = for τ N, H2 (+ ) = + and H 2 ( ) s strctly ncreasng n (N, ). Then, there s the unque postve t 2 such that H 2 (t 2 ) = T 2. () It s clear that H (, t 2 ) s contnuous and ncreasng n (t 2, ), H (, t 2 ) = + and H (t 2, t 2 ) = H 2 (t 2 ) = T 2 T. So, there s the unque postve t such that H (t, t 2 ) = T. (2) So, the system () has the unque soluton (t, t 2 ). Thus, (9) also has the unque soluton and t can be found by (). Ths completes the proof of Lemma 3. Lemmas 3 mply the followng man result. Theorem 5. The symmetrc water fllng game has the unque Nash equlbrum (T (ω, ω 2 ), T 2 (ω, ω 2 )) for g (, ), ω, ω 2 can be fount through t and t 2 from (7) whch are the unque soluton of the trangular system of equatons (). The assumpton that g < s essental for the unqueness of Nash Equlbrum as t s shown n the followng Proposton. Proposton. For g = the symmetrc water fllng game has a contnuum of Nash equlbra. 5. CONVERGENCE OF AN IWFA In ths secton we descrbe a verson of the water-fllng algorthm for fndng the Nash Equlbrum and supply a smple proof of ts convergence based on some monotonsty propertes. It s clear that H (ω, ω 2 ) =, 2 have the followng propertes, collected n the next Lemma, whch follow drectly from the explct formulas of the Nash Equlbrum. Lemma 4. () H (ω, ω 2 ), =, 2 are nonnegatve and contnuous, () H (ω, ω 2 ) s non-ncreasng on ω, () H (ω, ω 2 ) for ω, (v) H (ω, ω 2 ) = for enough bg ω, say for ω /N, (v) H (ω, ω 2 ) s non-decreasng by ω j (j ). These propertes gve a smple proof of the convergence of the followng water-fllng algorthm for fndng the Nash Equlbrum. Let ω and ω 2 be such that H (ω, ω 2 ) = H 2 (ω, ω 2 ) =, for example ω = ω 2 = /N. Let ω 2 = ω 2 and defne ω such that H (ω, ω 2 ) = T. Such ω exsts by Lemma 4()- (). Then, by Lemma 4(),(v) H 2 (ω, ω 2 ) =. Let ω 2 = ω and defne ω 2 2 such that H 2 (ω 2, ω 2 2) = T 2. Then, by Lemma 4(v) H (ω 2, ω 2 2) T and so on. So we have nonncreasng postve sequence (ω k, ω 2 k). Thus, t converges to an (ω, ω 2 ) whch produces a Nash Equlbrum.

5 6. CLOSED FORM SOLUTION FOR SYM- METRIC WATER FILLING GAME In ths secton, based on the proof of Lemma 3, we propose the soluton of the two players symmetrc water fllng game n the closed form. Wthout lost of generalty we can assume that T > T 2. Let k 2 be such that N k 2 + t 2 > N k 2. Then, snce H 2 (t 2 ) = T 2, we have that t 2 = ( + g) T 2 + P k 2 = πn P k2 = π. (3) Snce H 2 ( ) s strctly ncreasng, k 2 can be found from the condton H 2 (N k 2) < T 2 H 2 (N k 2 +). Hence, k 2 can be found from the followng equvalent condtons: ϕ 2 k 2 < T 2 ϕ 2 k 2 +, (4) ϕ 2 k = + g k π (Nk N ), = for k n, and ϕ 2 n+ =. Snce t s the root of the equaton H (, t 2 ) = T there s k k 2 such that N k + t > N k. So, () f k > k 2 then t = T + k =k 2 + () f k = k 2 then t = π N + + g π (gt 2 + N ) k 2 =, (5) k π = T + + g π (gt 2 + N ) k 2 =. (6) k π = Thus, k k 2 can be found as follows: () k = k 2 f T ϕ k 2 +, () otherwse k s gven by the condton: ϕ k = k =k g ϕ k < T ϕ k +, (7) π (N k N ) k 2 = π `( + g)n k N gt 2 for k [k 2 +, n], and ϕ n+ =. We can summarze the obtaned results n the followng theorem. Theorem 6. Let T > T 2. Then, the Nash equlbrum strateges are gven by >< t T gt2 + N + g f [, k 2 ], = t N f [k 2 +, k ], f [k +, n], () ( T 2 = + g (t2 N ) f [, k 2 ], f [k 2 +, n], k 2, t 2, k and t are gven by (4), (3), (7) and (5). 7. NUMERICALEAMPLE Let us demonstrate the closed form approach by a numercal example. Take n = 5, g =.9, T = 5, T 2 =.5, N = κ, κ =.7, π = /5 for [, 5]. Then, as the frst step we calculate ϕ 2 t for t [, 5]. In our case we get (,.74,.324,.963, 2.4). Then, by (4), k 2 = 3. Thus, by (3) t 2 = Then we calculate ϕ t for t [4, 5]. In our case we get (.3,4.3). So, by (7), k = 5. Usng (5), we fnd t = Thus, by () we have the followng equlbrum strateges T = (7.62, 6.694, 6.67, 4.3,.69) and T 2 = (.2,.99,.293,, ) wth payoffs.99 and.62. We have run IWFA, whch produced the same values for the optmal strateges and payoffs. However, we have observed that the convergence of IWFA s slow when g. In Fgure we have plotted the total error n strateges T k T 2 + T 2 k T 2 2, T k are the strateges produced by IWFA on the k-th teraton and T are the Nash equlbrum strateges. Our approach nstantaneously fnds the Nash equlbrum for all values of g. Also, t s nterestng to note that by () the quantty of channels as well as the channels themselves used by weaker player (wth smaller resources) s ndependent on behavor of the stronger player (wth bgger resources) but of course each player allocatng hs/her resources among these channels take nto account the opponent behavour. In Fgure 2, we compare the non-cooperatve approach wth the cooperatve approach. Specfcally, we compare the transmsson rates and ther sum under Nash equlbrum strateges and under strateges obtaned from the centralzed optmzaton of the sum of transmtters rates. The man conclusons are: the cost of anarchy s nearly zero for g [, /4] and then t grows up to 22% when g grows from /4 to ; the transmtter wth more resources gans sgnfcantly more from the centralzed optmzaton. Hence, the non-cooperatve approach results n a more far resource dstrbuton. In Table we gve strateges of both users obtaned n the case of the centralzed optmzaton for dfferent values of the crosstalk coeffcent g. Frst, we observe that when the crosstalk coeffcent s large, the users occupy dfferent resources. The user wth the larger average power takes better resources. When the crosstalk coeffcent s below.7, the users start to share the resources. As the value of the crosstalk coeffcent decreases, the 2nd user wth the smaller average power begns to occupy better resources. As expected, when the crosstalk coeffcent s very small, the optmal strateges start to look lke strateges whch are optmal n the case of no nterference.

6 error g=.9 g=.99 g= number of teratons g Table : Centralzed optmzaton Users strateges st user nd user st user nd user st user nd user st user nd user st user nd user st user nd user st user nd user Fgure : Convergence of IWFA centralzed wth respect to farness. We expect that the present approach can be generalzed to the case of more than two users. A detal analytcal study of the centralzed optmzaton s another future research topc Rate of Transmtter (Game) Rate of Transmtter 2 (Game) Sum of Rates (Game) Sum of Rates (Optm.) Rate of Transmtter (Optm.) Rate of Transmtter 2 (Optm.) g Fgure 2: Centralzed Optmzaton vs. Game. CONCLUSION We have consdered power control for wreless networks n optmzaton and game frameworks. Closed form solutons for the water fllng optmzaton problem and two players symmetrc water fllng games have been provded. Namely, now one can calculate optmal/equlbrum strateges wth a fnte number of arthmetc operatons. We have also provded a smple alternatve proof of convergence for a verson of teratve water fllng algorthm. It had been known before that the teratve water fllng algorthm converges very slow when the crosstalk coeffcent s close to one. For our closed form approach possble proxmty of the crosstalk coeffcent to one s not a problem. We have shown that when the crosstalk coeffcent s equal to one, there s a contnuum of Nash equlbra. Fnally, we have demonstarted that the prce of anarchy s small when the crosstalk coeffcent s small and that the decentralzed soluton s better than 9. APPENDI Proof of Theorem. Let T = (T,..., T n) be the optmal strategy. Snce P n = πt = T and T for [, n] there s a m [, n] such that Tm >. Let ǫ be any small enough postve number and k m. Let T ǫ,k = (T ǫ,k,..., Tn ǫ,k ) be such that >< Tm ǫ/π m for = m, T ǫ,k = Tk + ǫ/π k for = k, for {m, k}. T It s cleat that T ǫ,k also s a strategy for any enough small postve ǫ. Then, snce T s the optmal strategy we have that v(t ) v(t ǫ,k ). Thus, π k ln + T k N k + π m ln + T m π k ln + T k ǫ/π k Nk So, puttng ǫ we have that N m + π m ln + T m + ǫ/π m. Nm Tm + Nm Tk + for any m k. N k Thus, there s a postve ω such that ( = ω, for T >, T + N ω, for T =. So, the optmal strategy T s of the form T(ω) = (T (ω),..., T n(ω)) T (ω) = [/ω N ] + for [, n].

7 Remndng that the optmal strategy s non-negatve vector satsfyng the condton P n = πt = T we obtan that ω has to be found as a soluton of the equaton H T(ω) = T. It s clear that H T(+) = +, H T( ) s contnuous n (, ), H T(ω) = for ω [max (/N ), ) and H T( ) s strctly decreasng n (, max (/N )). Thus, there s unque postve ω such that H T(ω) = T. Ths completes the proof of Theorem. Proof of Theorem 2. Frst note H(ω) = for ω /N, H(ω) s strctly postve and decreasng n (, /N ). Let k [, n] be such that N k > ω, Nk+ Nn+ =. Then, ˆ/ω N = + /ω N for [, k] and ˆ/ω N = [k +, n]. So, + H(ω ) = k π (/ω N ). = Snce H(ω ) = T we have that ω = T + k = π. (9) k π N Because H s strctly decreasng on (, /N ) we can fnd k from the followng condton Snce = H(/N k) < T H(/N k+). k k+ π (Nk+ N ) = π (Nk+ N ), = the nteger k can be found from the followng equvalent condton ϕ t = = ϕ k < T ϕ k+, (2) t π (Nt N ) for t [, n]. = Therefore, Theorem, (9) and (2) mply Theorem 2. Proof of Theorem 3. The Lagrangan correspondng to mnmzaton of v j subject to the constrant (4) and nonnegatvty constrants on T j s gven by! L j T j n! k = π k ln + + ω j π k T j k T j k= gt m k + N k +ν j ( T j ), k= wth m j. Dfferentatng the Lagrangan wth respect to T j and equatng the dervatve to zero, we obtan gt m + T j + N + νj π = ω j. (2) Now, usng the complmentary slackness condton ν j T j =, we obtan condton (5). Snce the left hand sde of equaton (2) s postve, the Lagrange multpler ω j s postve as well. Proof of Lemma. () follows drectly from (5) = T 2 =. T () Let T Thus, that Thus, > and T 2 T =. Then by (5) we have that + N = ω. ω > N and T = ω N. Then, by (5) we have ω 2 gt and the result follows. + N = g(. ω N ) + N g( ω N ) ω 2 N () can be proved smlarly to (). (v) Let T > and T 2 >. Then, by (5)we have that ω > N and ω 2 > N. Also, by (5) we have that T and T 2 are gven by (6). Then, snce T > and T 2 > we have that I (ω, ω 2 ). Ths completes the proof of the lemma. Proof of Lemma 2. Let, for example, ω < ω 2. It s clear that I (ω, ω 2 ) = I (ω, ω 2 ). Let I (ω, ω 2 ). Then ether /ω 2 N < /ω or /ω2 g/µ g N < mn{/ω, /ω 2 }. Snce ω < ω 2, mn{/ω, /ω 2 } = /ω 2 and /ω2 g/ω g < /ω 2. Thus, I (ω, ω 2 ) = I (ω, ω 2 ). Let I (ω, ω 2 ). Then ether /ω N < /ω 2 or /ω g/ω 2 g N < mn{/ω, /ω 2 }. Snce ω < ω 2, then /ω g/ω 2 g > /ω 2. So, I (ω, ω 2 ) = = I (ω, ω 2 ). Let I (ω, ω 2 ). Then N < mn{ /ω g/ω 2 g, /ω2 g/ω g } = /ω2 g/ω g. Thus, I (ω, ω 2 ) = I (ω, ω 2 ). Ths completes the proof of Lemma 2. Proof of Corollary. Assume that ω ω 2. Then /ω g/ω 2 /ω 2 g/ω. Thus, T (ω, ω 2 ) T 2 (ω, ω 2 ) for [, n]. So, T = π T (ω, ω 2 ) π T 2 (ω, ω 2 ) = T 2. = = Ths contradcton completes the proof of Corollary. Proof of Proposton. Suppose that (T, T 2 ) be a Nash equlbrum. Then, smlarly to Lemma, we have to consder three cases ()-() at least one of components of the vector (T, T 2 ) s postve.

8 () Let T > and T 2 =. Then, by (5), we have that Thus, Then, by (5) ω 2 () Let T 2 have that and () Let T T + N = ω. ω > N and T = ω N. T + N = > and T T 2 ω N + N = ω. =. Then, smlarly to (), we = ω 2 N ω 2 > N, ω ω 2. > and T 2 T >. Then, by (5), have that + T 2 + N = ω = ω 2. Assume that ω > ω 2 then () does not hold, so T = for each whch contradcts to (4). Smlarly, the case ω < ω 2 cannot hold. Thus, ω = ω 2 = ω. So, T be any non-negatve such that and = T + T 2 = [/ω N ] + π T = T, = and T 2, [, n] have to π T 2 = T 2, ω s the unque postve root of the equaton π [/ω N ] + = T + T 2. = It s clear that there s a contnuum of such strateges. For example f (T, T 2 ) s the one of them, and let Tk, Tk 2 > and Tm, Tm 2 > for some k and m. Then, t s clear that the followng strateges for any enough small postve ǫ are also optmal: >< T for k, m, T = T + ǫ for = k, T ǫπ k /π m for = m, >< T 2 for k, m, T 2 = T 2 ǫ for = k, T 2 + ǫπ k /π m for = m. Ths completes the proof of Proposton. [2] E. Altman, K. Avrachenkov, G. Mller and B. Prabhu, Dscrete power control: cooperatve and non-cooperatve optmzaton, n Proceedngs of IEEE INFOCOM 27. An extended verson s avalable as INRIA Research Report no.5. [3] S.T. Chung and J.M. Coff, Rate and power control n a two-user multcarrer channel wth no coordnaton: the optmal scheme vs. suboptmal methods, n Proceedngs of IEEE VTC 2-Fall, v.3, pp , 22. [4] T. Cover and J. Thomas, Elements of Informaton Theory, Wley, 99. [5] W. R. Henzelman, A. Chandrakasan, and H. Balakrshnan, Energy-effcent communcaton protocol for wreless mcrosensor networks, n Proc. of the 33rd Annual Hawa Internatonal Conference on System Scences, v.2, Jan. 2. [6] A.J. Goldsmth and P.P. Varaya, Capacty of fadng channels wth channel sde nformaton, IEEE Trans. Informaton Theory, v.43(6), pp , 997. [7] T. J. Kwon and M. Gerla, Clusterng wth power control, n Proc. IEEE Mltary Communcatons Conference (MILCOM 99), v.2, Atlantc Cty, NJ, USA, 999, pp [] L. La and H. El Gamal, The water-fllng game n fadng multple access channels, submtted to IEEE Trans. Informaton Theory, November 25, avalable at helgamal/. [9] C. R. Ln and M. Gerla, Adaptve clusterng for moble wreless networks, IEEE JSAC, v.5, no.7, pp , 997. [] Z.-Q. Luo and J.-S. Pang, Analyss of teratve waterfllng algorthm for multuser power control n dgtal subscrber lnes, EURASIP Journal on Appled Sgnal Processng, 26. [] O. Popescu and C. Rose, Water fllng may not good neghbors make, n Proceedngs of GLOBECOM 23, v.3, pp , 23. [2] D.C. Popescu, O. Popescu and C. Rose, Interference avodance versus teratve water fllng n multaccess vector channels, n Proceedngs of IEEE VTC 24 Fall, v.3, pp , 24. [3] D. Tse and P. Vswanath, Fundamentals of Wreless Communcaton, Cambrdge Unversty Press, 25. [4] W. Yu, Competton and cooperaton n mult-user communcaton envronements, PhD Thess, Stanford Unversty, June 22. [5] W. Yu, G. Gns and J.M. Coff, Dstrbuted multuser power control for dgtal subscrber lnes, IEEE JSAC, v.2, pp.5 5, 22.. REFERENCES [] E. Altman, K. Avrachenkov, and A. Garnaev, Jammng game n wreless networks wth transmsson cost. Lecture Notes n Computer Scence, v.4465, pp.-2, 27.

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

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

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

Computing Correlated Equilibria in Multi-Player Games

Computing Correlated Equilibria in Multi-Player Games Computng Correlated Equlbra n Mult-Player Games Chrstos H. Papadmtrou Presented by Zhanxang Huang December 7th, 2005 1 The Author Dr. Chrstos H. Papadmtrou CS professor at UC Berkley (taught at Harvard,

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

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

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

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

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

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

CS286r Assign One. Answer Key

CS286r Assign One. Answer Key CS286r Assgn One Answer Key 1 Game theory 1.1 1.1.1 Let off-equlbrum strateges also be that people contnue to play n Nash equlbrum. Devatng from any Nash equlbrum s a weakly domnated strategy. That s,

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

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 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

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

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

Games of Threats. Elon Kohlberg Abraham Neyman. Working Paper

Games of Threats. Elon Kohlberg Abraham Neyman. Working Paper Games of Threats Elon Kohlberg Abraham Neyman Workng Paper 18-023 Games of Threats Elon Kohlberg Harvard Busness School Abraham Neyman The Hebrew Unversty of Jerusalem Workng Paper 18-023 Copyrght 2017

More information

CHAPTER-5 INFORMATION MEASURE OF FUZZY MATRIX AND FUZZY BINARY RELATION

CHAPTER-5 INFORMATION MEASURE OF FUZZY MATRIX AND FUZZY BINARY RELATION CAPTER- INFORMATION MEASURE OF FUZZY MATRI AN FUZZY BINARY RELATION Introducton The basc concept of the fuzz matr theor s ver smple and can be appled to socal and natural stuatons A branch of fuzz matr

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

A Bayesian Jamming Game in an OFDM Wireless Network

A Bayesian Jamming Game in an OFDM Wireless Network A Bayesan Jammng Game n an OFDM Wreless Networ Andrey Garnaev Yezeael Hayel Etan Altman To cte ths verson: Andrey Garnaev Yezeael Hayel Etan Altman. A Bayesan Jammng Game n an OFDM Wreless Networ. WOpt

More information

The Second Anti-Mathima on Game Theory

The Second Anti-Mathima on Game Theory The Second Ant-Mathma on Game Theory Ath. Kehagas December 1 2006 1 Introducton In ths note we wll examne the noton of game equlbrum for three types of games 1. 2-player 2-acton zero-sum games 2. 2-player

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

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

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

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

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

The Order Relation and Trace Inequalities for. Hermitian Operators

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

More information

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

Lecture 21: Numerical methods for pricing American type derivatives

Lecture 21: Numerical methods for pricing American type derivatives Lecture 21: Numercal methods for prcng Amercan type dervatves Xaoguang Wang STAT 598W Aprl 10th, 2014 (STAT 598W) Lecture 21 1 / 26 Outlne 1 Fnte Dfference Method Explct Method Penalty Method (STAT 598W)

More information

Complete subgraphs in multipartite graphs

Complete subgraphs in multipartite graphs Complete subgraphs n multpartte graphs FLORIAN PFENDER Unverstät Rostock, Insttut für Mathematk D-18057 Rostock, Germany Floran.Pfender@un-rostock.de Abstract Turán s Theorem states that every graph G

More information

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

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

More information

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

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

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

Stanford University CS359G: Graph Partitioning and Expanders Handout 4 Luca Trevisan January 13, 2011

Stanford University CS359G: Graph Partitioning and Expanders Handout 4 Luca Trevisan January 13, 2011 Stanford Unversty CS359G: Graph Parttonng and Expanders Handout 4 Luca Trevsan January 3, 0 Lecture 4 In whch we prove the dffcult drecton of Cheeger s nequalty. As n the past lectures, consder an undrected

More information

A new Approach for Solving Linear Ordinary Differential Equations

A new Approach for Solving Linear Ordinary Differential Equations , ISSN 974-57X (Onlne), ISSN 974-5718 (Prnt), Vol. ; Issue No. 1; Year 14, Copyrght 13-14 by CESER PUBLICATIONS A new Approach for Solvng Lnear Ordnary Dfferental Equatons Fawz Abdelwahd Department of

More 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

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

APPENDIX A Some Linear Algebra

APPENDIX A Some Linear Algebra APPENDIX A Some Lnear Algebra The collecton of m, n matrces A.1 Matrces a 1,1,..., a 1,n A = a m,1,..., a m,n wth real elements a,j s denoted by R m,n. If n = 1 then A s called a column vector. Smlarly,

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

PROBLEM SET 7 GENERAL EQUILIBRIUM

PROBLEM SET 7 GENERAL EQUILIBRIUM PROBLEM SET 7 GENERAL EQUILIBRIUM Queston a Defnton: An Arrow-Debreu Compettve Equlbrum s a vector of prces {p t } and allocatons {c t, c 2 t } whch satsfes ( Gven {p t }, c t maxmzes βt ln c t subject

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

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

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

(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

Lectures - Week 4 Matrix norms, Conditioning, Vector Spaces, Linear Independence, Spanning sets and Basis, Null space and Range of a Matrix

Lectures - Week 4 Matrix norms, Conditioning, Vector Spaces, Linear Independence, Spanning sets and Basis, Null space and Range of a Matrix Lectures - Week 4 Matrx norms, Condtonng, Vector Spaces, Lnear Independence, Spannng sets and Bass, Null space and Range of a Matrx Matrx Norms Now we turn to assocatng a number to each matrx. We could

More 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

The Expectation-Maximization Algorithm

The Expectation-Maximization Algorithm The Expectaton-Maxmaton Algorthm Charles Elan elan@cs.ucsd.edu November 16, 2007 Ths chapter explans the EM algorthm at multple levels of generalty. Secton 1 gves the standard hgh-level verson of the algorthm.

More information

Lecture 4. Instructor: Haipeng Luo

Lecture 4. Instructor: Haipeng Luo Lecture 4 Instructor: Hapeng Luo In the followng lectures, we focus on the expert problem and study more adaptve algorthms. Although Hedge s proven to be worst-case optmal, one may wonder how well t would

More information

Thermodynamics and statistical mechanics in materials modelling II

Thermodynamics and statistical mechanics in materials modelling II Course MP3 Lecture 8/11/006 (JAE) Course MP3 Lecture 8/11/006 Thermodynamcs and statstcal mechancs n materals modellng II A bref résumé of the physcal concepts used n materals modellng Dr James Ellott.1

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

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

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

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

Lecture 5 Decoding Binary BCH Codes

Lecture 5 Decoding Binary BCH Codes Lecture 5 Decodng Bnary BCH Codes In ths class, we wll ntroduce dfferent methods for decodng BCH codes 51 Decodng the [15, 7, 5] 2 -BCH Code Consder the [15, 7, 5] 2 -code C we ntroduced n the last lecture

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

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

Inner Product. Euclidean Space. Orthonormal Basis. Orthogonal

Inner Product. Euclidean Space. Orthonormal Basis. Orthogonal Inner Product Defnton 1 () A Eucldean space s a fnte-dmensonal vector space over the reals R, wth an nner product,. Defnton 2 (Inner Product) An nner product, on a real vector space X s a symmetrc, blnear,

More information

COMPLEX NUMBERS AND QUADRATIC EQUATIONS

COMPLEX NUMBERS AND QUADRATIC EQUATIONS COMPLEX NUMBERS AND QUADRATIC EQUATIONS INTRODUCTION We know that x 0 for all x R e the square of a real number (whether postve, negatve or ero) s non-negatve Hence the equatons x, x, x + 7 0 etc are not

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

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

Section 8.3 Polar Form of Complex Numbers

Section 8.3 Polar Form of Complex Numbers 80 Chapter 8 Secton 8 Polar Form of Complex Numbers From prevous classes, you may have encountered magnary numbers the square roots of negatve numbers and, more generally, complex numbers whch are the

More information

On Information Theoretic Games for Interference Networks

On Information Theoretic Games for Interference Networks On Informaton Theoretc Games for Interference Networks Suvarup Saha and Randall A. Berry Dept. of EECS Northwestern Unversty e-mal: suvarups@u.northwestern.edu rberry@eecs.northwestern.edu Abstract The

More information

Randić Energy and Randić Estrada Index of a Graph

Randić Energy and Randić Estrada Index of a Graph EUROPEAN JOURNAL OF PURE AND APPLIED MATHEMATICS Vol. 5, No., 202, 88-96 ISSN 307-5543 www.ejpam.com SPECIAL ISSUE FOR THE INTERNATIONAL CONFERENCE ON APPLIED ANALYSIS AND ALGEBRA 29 JUNE -02JULY 20, ISTANBUL

More information

Hila Etzion. Min-Seok Pang

Hila Etzion. Min-Seok Pang RESERCH RTICLE COPLEENTRY ONLINE SERVICES IN COPETITIVE RKETS: INTINING PROFITILITY IN THE PRESENCE OF NETWORK EFFECTS Hla Etzon Department of Technology and Operatons, Stephen. Ross School of usness,

More information

Market structure and Innovation

Market structure and Innovation Market structure and Innovaton Ths presentaton s based on the paper Market structure and Innovaton authored by Glenn C. Loury, publshed n The Quarterly Journal of Economcs, Vol. 93, No.3 ( Aug 1979) I.

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

Canonical transformations

Canonical transformations Canoncal transformatons November 23, 2014 Recall that we have defned a symplectc transformaton to be any lnear transformaton M A B leavng the symplectc form nvarant, Ω AB M A CM B DΩ CD Coordnate transformatons,

More information

Lecture 17 : Stochastic Processes II

Lecture 17 : Stochastic Processes II : Stochastc Processes II 1 Contnuous-tme stochastc process So far we have studed dscrete-tme stochastc processes. We studed the concept of Makov chans and martngales, tme seres analyss, and regresson analyss

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

Perron Vectors of an Irreducible Nonnegative Interval Matrix

Perron Vectors of an Irreducible Nonnegative Interval Matrix Perron Vectors of an Irreducble Nonnegatve Interval Matrx Jr Rohn August 4 2005 Abstract As s well known an rreducble nonnegatve matrx possesses a unquely determned Perron vector. As the man result of

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

Infinitely Split Nash Equilibrium Problems in Repeated Games

Infinitely Split Nash Equilibrium Problems in Repeated Games Infntely Splt ash Equlbrum Problems n Repeated Games Jnlu L Department of Mathematcs Shawnee State Unversty Portsmouth, Oho 4566 USA Abstract In ths paper, we ntroduce the concept of nfntely splt ash equlbrum

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

Appendix B. The Finite Difference Scheme

Appendix B. The Finite Difference Scheme 140 APPENDIXES Appendx B. The Fnte Dfference Scheme In ths appendx we present numercal technques whch are used to approxmate solutons of system 3.1 3.3. A comprehensve treatment of theoretcal and mplementaton

More 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

U.C. Berkeley CS294: Spectral Methods and Expanders Handout 8 Luca Trevisan February 17, 2016

U.C. Berkeley CS294: Spectral Methods and Expanders Handout 8 Luca Trevisan February 17, 2016 U.C. Berkeley CS94: Spectral Methods and Expanders Handout 8 Luca Trevsan February 7, 06 Lecture 8: Spectral Algorthms Wrap-up In whch we talk about even more generalzatons of Cheeger s nequaltes, and

More information

Entropy of Markov Information Sources and Capacity of Discrete Input Constrained Channels (from Immink, Coding Techniques for Digital Recorders)

Entropy of Markov Information Sources and Capacity of Discrete Input Constrained Channels (from Immink, Coding Techniques for Digital Recorders) Entropy of Marov Informaton Sources and Capacty of Dscrete Input Constraned Channels (from Immn, Codng Technques for Dgtal Recorders). Entropy of Marov Chans We have already ntroduced the noton of entropy

More information

A Local Variational Problem of Second Order for a Class of Optimal Control Problems with Nonsmooth Objective Function

A Local Variational Problem of Second Order for a Class of Optimal Control Problems with Nonsmooth Objective Function A Local Varatonal Problem of Second Order for a Class of Optmal Control Problems wth Nonsmooth Objectve Functon Alexander P. Afanasev Insttute for Informaton Transmsson Problems, Russan Academy of Scences,

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

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

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

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

a b a In case b 0, a being divisible by b is the same as to say that

a b a In case b 0, a being divisible by b is the same as to say that Secton 6.2 Dvsblty among the ntegers An nteger a ε s dvsble by b ε f there s an nteger c ε such that a = bc. Note that s dvsble by any nteger b, snce = b. On the other hand, a s dvsble by only f a = :

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

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

Linear Algebra and its Applications

Linear Algebra and its Applications Lnear Algebra and ts Applcatons 4 (00) 5 56 Contents lsts avalable at ScenceDrect Lnear Algebra and ts Applcatons journal homepage: wwwelsevercom/locate/laa Notes on Hlbert and Cauchy matrces Mroslav Fedler

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

LECTURE 9 CANONICAL CORRELATION ANALYSIS

LECTURE 9 CANONICAL CORRELATION ANALYSIS LECURE 9 CANONICAL CORRELAION ANALYSIS Introducton he concept of canoncal correlaton arses when we want to quantfy the assocatons between two sets of varables. For example, suppose that the frst set of

More information

Edge Isoperimetric Inequalities

Edge Isoperimetric Inequalities November 7, 2005 Ross M. Rchardson Edge Isopermetrc Inequaltes 1 Four Questons Recall that n the last lecture we looked at the problem of sopermetrc nequaltes n the hypercube, Q n. Our noton of boundary

More information

General viscosity iterative method for a sequence of quasi-nonexpansive mappings

General viscosity iterative method for a sequence of quasi-nonexpansive mappings Avalable onlne at www.tjnsa.com J. Nonlnear Sc. Appl. 9 (2016), 5672 5682 Research Artcle General vscosty teratve method for a sequence of quas-nonexpansve mappngs Cuje Zhang, Ynan Wang College of Scence,

More information

Computing a Cournot Equilibrium in Integers

Computing a Cournot Equilibrium in Integers Computng a Cournot Equlbrum n Integers Mchael J. Todd December 6, 2013 Abstract We gve an effcent algorthm for computng a Cournot equlbrum when the producers are confned to ntegers, the nverse demand functon

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

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

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

More information

Example: (13320, 22140) =? Solution #1: The divisors of are 1, 2, 3, 4, 5, 6, 9, 10, 12, 15, 18, 20, 27, 30, 36, 41,

Example: (13320, 22140) =? Solution #1: The divisors of are 1, 2, 3, 4, 5, 6, 9, 10, 12, 15, 18, 20, 27, 30, 36, 41, The greatest common dvsor of two ntegers a and b (not both zero) s the largest nteger whch s a common factor of both a and b. We denote ths number by gcd(a, b), or smply (a, b) when there s no confuson

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

VARIATION OF CONSTANT SUM CONSTRAINT FOR INTEGER MODEL WITH NON UNIFORM VARIABLES

VARIATION OF CONSTANT SUM CONSTRAINT FOR INTEGER MODEL WITH NON UNIFORM VARIABLES VARIATION OF CONSTANT SUM CONSTRAINT FOR INTEGER MODEL WITH NON UNIFORM VARIABLES BÂRZĂ, Slvu Faculty of Mathematcs-Informatcs Spru Haret Unversty barza_slvu@yahoo.com Abstract Ths paper wants to contnue

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

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