Joint Virtual Machine and Bandwidth Allocation in Software Defined Network (SDN) and Cloud Computing Environments

Size: px
Start display at page:

Download "Joint Virtual Machine and Bandwidth Allocation in Software Defined Network (SDN) and Cloud Computing Environments"

Transcription

1 IEEE ICC Next-Geneaton Netwokng Symposum 1 Jont Vtual Machne and Bandwdth Allocaton n Softwae Defned Netwok (SDN) and Cloud Computng Envonments Jonathan Chase, Rakpong Kaewpuang, Wen Yonggang, and Dust Nyato School of Compute Engneeng, Nanyang Technologcal Unvesty, Sngapoe. Abstact Cloud computng povdes uses wth geat flexblty when povsonng esouces, wth cloud povdes offeng a choce of esevaton and on-demand puchasng optons. Resevaton plans offe cheape pces, but must be chosen n advance, and theefoe must be appopate to uses equements. If demand s uncetan, the esevaton plan may not be suffcent and on-demand esouces have to be povsoned. Pevous wok focused on optmally placng vtual machnes wth cloud povdes to mnmze total cost. Howeve, many applcatons eque lage amounts of netwok bandwdth. Theefoe, consdeng only vtual machnes offes an ncomplete vew of the system. Explotng ecent developments n softwae defned netwokng (SDN), we popose a unfed appoach that ntegates vtual machne and netwok bandwdth povsonng. We solve a stochastc ntege pogammng poblem to obtan an optmal povsonng of both vtual machnes and netwok bandwdth, when demand s uncetan. Numecal esults clealy show that ou poposed soluton mnmzes uses costs and povdes supeo pefomance to altenatve methods. We beleve that ths ntegated appoach s the way fowad fo cloud computng to suppot netwok ntensve applcatons. Index Tems Cloud computng, softwae defned netwok, vtual machne, bandwdth allocaton I. INTRODUCTION Cloud computng pemts uses to access computng esouces fom cloud povdes ove the Intenet. Vtualzaton technology allows uses to access computng esouces by entng vtual machnes (VMs) that ae taloed to the equements. Cloud povdes offe pepad esevaton nstances at a educed usage pce, wth addtonal VMs obtanable on demand at a hghe pce. When demand s unknown n advance, thee s a hgh sk of ethe undepovsonng (when the use eseves too few esouces fo the needs) o ovepovsonng (use eseves too many esouces and pays fo esouces they do not need). A boke can be used to mnmze cost though effcent VM povsonng. Howeve, wth many cloud-based applcatons equng a lage amount of Intenet bandwdth, an appoach that focuses puely on VM placement s too smplstc, a unfed appoach that ncopoates bandwdth esevaton s needed. To acheve ths, we explot the contol offeed by a new and developng technology called softwae defned netwokng (SDN) [1]. SDN s an appoach to netwok vtualzaton that sepaates netwok contol fom the mechancs of outng. SDN uses a thee laye achtectue wth an applcaton laye whee uses can un contol pogams that ae mplemented tanspaently by the lowe level contol and nfastuctue layes. Ths applcaton laye allows the netwok contolle to offe a vtual netwok, wth vtual outes and lnks, that the contolle then mplements on the physcal netwok. Thus netwok contolles can offe esevaton and on-demand bandwdth. In ths pape, we popose a unfed optmal povsonng algothm that places VMs and netwok bandwdth to mnmze a uses costs. The algothm makes a decson to eseve VMs and netwok bandwdth fom cetan cloud povdes to mnmze costs, by tadng off the cost of ovepovsonng aganst the cost of on-demand esouces. We fomulate a stochastc ntege pogammng (SIP) poblem wth two-stage ecouse to obtan the optmal decson. We demonstate the soluton s effectveness though vaous numecal esults and show that uses costs can be mnmzed whle meetng use demand and cloud povde capacty lmts. II. RELATED WORK Cloud computng typcally offes thee man sevce models, Softwae-as-a-Sevce (SaaS), Platfom-as-a-Sevce (PaaS), and Infastuctue-as-a-Sevce (IaaS). IaaS sevces, such as Amazon EC2 [2], explot vtualzaton technology to offe esouces to uses n the fom of VMs. Allocaton methods may take a heustc appoach, such as n [3], whch also suppots mgaton. Altenatvely, [4] descbes an exact povsonng algothm that accommodates both VM demand and pce uncetanty. Intoducng the use of two-stage stochastc optmzaton [5] to computng esouce povsonng, these woks fnd optmal solutons to the NP-had VM allocaton poblem. Smlaly, [6] fomulates mxed ntege poblems, and compaes them, focusng on mnmzng executon tme. The lage amounts of bandwdth equed by moden cloud applcatons can be guaanteed though bandwdth esevaton. Fo example, [7] addesses the poblem of mappng use equests to data centes and outng the esponses but does not consde computaton. The OpenFlow [8] mplementaton of SDN allows the flexble allocaton of bandwdth though vtual netwoks, and theefoe the esevaton of bandwdth. Combnng ths powe wth methods used n Vtual Netwok Embeddng (VNE), bandwdth allocaton can be pefomed fo the cloud. Fo example, [9] fomulates the jont poblem of vtual node and lnk allocaton as an Intege Pogammng poblem, although wthout a stochastc element. Lke VM allocaton, ths poblem s NP-had, so s often solved wth heustcs nstead [10] /14/$ IEEE 2969

2 IEEE ICC Next-Geneaton Netwokng Symposum 2 SDN Contolle Bandwdth Povsonng VM Povsonng Demand Requests R1 R5 P1 Fg. 1. Tmelne showng the stages of decson-makng. Resevaton decsons last fo a longe peod of tme, wth second stage decsons made fo a shote duaton once demand s known. R2 R6 P2 R3 The above wok lays the foundaton fo a jont VM and bandwdth allocaton soluton. Howeve, as fa as we ae awae, thee ae cuently no woks that addess the poblem dectly. It s ou belef that the two must be consdeed togethe to acheve a globally optmal soluton to the cloud esouce povsonng poblem. Use R7 R4 R8 Te 1 Te 2 ISP Netwoks P3 P4 Cloud Povdes III. SYSTEM MODEL A. Cloud and Netwok Resouces The system model unde consdeaton fo jont VM and bandwdth allocaton s composed of cloud sevce povdes and ISPs. Thee s a cental contolle ecevng demand equests fo VM povsonng fom uses. The contolle then povsons the equed VMs fom cloud povdes and povsons vtualzed bandwdth fom ISPs. ISPs can mplement SDN (e.g. wth OpenFlow-enabled swtches), enablng the cental contolle to make bandwdth povsonng and outng decsons on vtualzed outes. A pactcal mplementaton may eque sepaate contolles to admnste netwoks and povdes, howeve, we ae nteested n the decson-makng pocess, whch we model wth a sngle contolle. The povsonng decson s made n two stages (Fg. 1). The fst stage s the esevaton phase, whee VMs and bandwdth ae eseved fo a long peod of tme, typcally a yea, n advance, befoe actual demand s known. The second stage happens at the tme of use, when the uses actual demand s ealzed (e.g., daly demand). The second stage s dvded nto two phases,.e., utlzaton and on-demand phases. In the utlzaton phase, the needed eseved esouces ae used, usually at a low pce. If the actual demand s hghe than the eseved esouces, the algothm entes the on-demand phase. In the on-demand phase, addtonal esouces can be povsoned at a hghe cost to satsfy any unsatsfed demand. Theefoe, t s mpotant fo the fst stage to be optmally detemned as t s sgnfcantly cheape than the second stage, but less flexble, due to the long duaton of esevaton. In the system shown n Fg. 2, an ISP manages a set of vtualzed outes, denoted by R = {1,...,R}, whee R s the total numbe of outes. Each oute has a bandwdth capacty, t. Resevaton, utlzaton, and on-demand bandwdth costs, though oute, ae gven by c (e), c, and c, espectvely. A set of cloud povdes s denoted by P = {1,...,P}, whee P s the total numbe of cloud povdes. Each cloud povde, p, has thee esouces,.e., pocessng powe, stoage, and ntenal netwok bandwdth, whose capactes ae gven by t (h) p, t (s) p, and t (n) p, espectvely. Hee ntenal bandwdth s Fg. 2. System model. Bandwdth esevaton s made fom ISP netwoks, VMs ae placed wth cloud povdes. Cental contolle makes povsonng choces based on use equements. bandwdth allocated fo a VM (e.g., fo a seve nsde a data cente), whle extenal bandwdth efes to the bandwdth of the outes of the ISP. V = {V 1,V 2,...,V last } denotes the set of VM classes. Each class has dstnct esouce equements, wth d (h), d (s), d (n), and d (b), denotng a VM fom class V s equements fo pocessng, stoage, ntenal bandwdth, and extenal bandwdth, espectvely. The costs of esevng, utlzng, and obtanng on-demand, a VM, fom class V, fom povde p, ae gven by c (e), c, and c, espectvely. All costs ae known n advance, but the VM demand (and theefoe the extenal bandwdth demand) s uncetan. B. Uncetanty of VM and Bandwdth Demand Wth uncetan demand, the numbe of VMs equed fom each class s unknown at the pont of esevaton. D = {d 1,d 2,...,d V } denotes the set of possble numbes of VMs that can be eseved fom class V. The set of all possble VM demand values can theefoe be gven by D = D = D 1 D 2 D V (1) whee and ae the Catesan poduct. Gven uncetan VM demand, the total bandwdth that needs to be obtaned s also unknown at the pont of esevaton. Snce each VM class,(1) can be used to gve the set of possble bandwdth equements fo class V. The set of possble bandwdth equements fo VM class V s defned as follows: has a fxed extenal bandwdth equement, d (b) B = D d (b). (2) Then the set of possble extenal bandwdth equements of all VM classes s expessed as follows: B = D d (b). (3) 2970

3 IEEE ICC Next-Geneaton Netwokng Symposum 3 IV. PROBLEM FORMULATION A. Stochastc Optmzaton Poblem If VM demand wee known at the pont of esevaton, the poblem would be a smple detemnstc optmzaton, wth no on-demand phase. Snce demand s uncetan, we need to apply a two-stage stochastc optmzaton. The stochastc optmzaton poblem can be fomulated as follows: ( mn c (e) (e),y (e) (e) + c (e) Y (e) R p P ) (4) +E[L( (e),y (e),ω)]. The objectve functon n (4) mnmzes the total cost of esevaton added to the expected cost fom the second stage. The decson vaables (e) and Y (e) denote the amount of bandwdth eseved fom oute and the numbe of VMs of class eseved fom cloud povde p, espectvely. The expected second stage cost s epesented by E[L( (e),y (e),ω)], whee ω denotes a scenao of demand: ω Ω = D B. Fo a gven ω, the second stage cost functon s gven n (5), whee decson vaables (ω) and (ω) denote the amount of povsoned bandwdth fom oute n the utlzaton and on-demand phases gven demand scenao ω, espectvely. Decson vaables Y (ω) and Y (ω) denote the numbe of VMs fom class povsoned n the utlzaton and on-demand phases fom cloud povde p, espectvely. The full fomulaton, theefoe, s to mnmze (4), subject to the followng constants: p P Y (e),y X (e) (ω),y (ω) {0, 1,...} (6), (ω),x (ω) 0 (7) (ω)+ (ω) t, R (8) d (h) d (s) d (n) Y (ω)+y (ω)) t(h) p,p P (9) (ω)+y (ω)) t(s) p,p P (10) (ω)+y (ω)) t(n) p,p P (11) (ω) (e), R (12) (e) (ω) Y,p P,V V (13) (ω)+y (ω)) v (ω), v V. (14) (6) pevents negatve o patal povsonng of a VM. Smlaly, (7) ensues non-negatve bandwdth povsonng, but pemts non-ntege amounts. (8)-(11) ensue that the povsoned esouces do not exceed capacty, wth t gvng the capacty of oute, and t p (n) gvng the capacty of cloud povde p fo each of the thee esouces,.e., pocessng powe, stoage, and ntenal bandwdth, espectvely. (12) and (13) ensues that the povsonng n utlzaton phase does not exceed the eseved amount. (14) ensues the VM demand ealzed n the second stage, denoted by v (ω) fo VM class, s satsfed., and t (h) p, t (s) p Bandwdth demand must also be met, as well as ensung that flow s conseved acoss the whole netwok. We consde the netwok of outes as a gaph, G =(R, E), whee each oute n R s a vetex, and each lnk n E s an edge, wth flow pesevaton defned n tems of edges. Howeve, as bandwdth s allocated at the oute, we defne a set of constants to contol flow based on outes as follows: R(p) k R out() R pov (ω)+ (ω) k R use R(p m) R out( m) (ω)+x k (ω)+x d (b) (ω) (ω) p P (ω)+y (ω)),p P (15) (ω)+ (ω), R d (b) (ω)+ (ω) R pov (ω)+ (ω) d (b) p p m (16) (ω)+y (ω)) (17) (ω)+ (ω) (18) (ω)+y (ω)), p m P adj (19) (ω)+ (ω) m (ω)+x (ω), m R adj. (20) The man dea of the above constants s to balance the nput bandwdth wth output bandwdth at the outes. The nequaltes ensue that the bandwdth supply s at least suffcent to meet demand thoughout the netwok, elyng on the mnmzaton to emove excess bandwdth povsonng. Total actual povsoned bandwdth s gven by the utlzaton and on-demand values fo a gven ealzaton. (15) ensues that the bandwdth equements of each ndvdual cloud povde ae met, whee R(p) s a set of outes dectly connected to cloud povde p. (16) ensues that the bandwdth allocated to each oute, R, s suppled suffcently by the outes, k R out (), that connect to t wth the outwad lnks. (17) ensues that the total bandwdth equements of all cloud povdes ae met by the outes that dectly 2971

4 IEEE ICC Next-Geneaton Netwokng Symposum 4 L( (e),y (e),ω)= mn (ω), (ω),y (ω),y c (ω) (ω)+c (ω)+c Y (ω)+c Y (ω) (5) connect to them. The set of all outes that lnk dectly to cloud povdes s gven by R pov. (18) s to ensue that the total bandwdth supply s povded to the use, whee R use s the set of all outes wth connectons dectly to the use. In (19) the set P adj s the set of dsjont pas, p m,of povdes whee the povdes n the pa ae connected dectly to at least one oute n common. R(p m ) denotes the set of outes that connect dectly to ethe of the povdes n the pa, p m. In (20), the set m R adj, s the set of dsjont pas of outes, m, whee the outes n the pa both eceve nput fom at least one oute n common, but ae not dectly connected to each othe. R out ( m ) denotes the set of all outes that connect dectly to the outes n pa m wth the outwad lnks. Whee outes have two o moe outwad lnks, thee s a dange that the same bandwdth allocaton wll be assumed to supply multle outes o cloud povdes; (19) and (20) pevent ths. Although the flow s symmetcal, we consde the lnks that come fom the use to be nwad, and the lnks that go towads the cloud povdes to be the outwad lnks. B. Detemnstc Equvalent Fomulaton The stochastc fomulaton above can be tansfomed nto a detemnstc equvalent fomulaton by ntoducng fou vaables as equvalents to the decson vaables n (4). These new vaables ae (b, d), (b, d), Y (b, d), and Y (b, d). When VM and bandwdth demand s ealzed, we have the obseved values d Dand b B. Ths pemts the efomulaton of the objectve functon defned n (4) as shown n (21). In ths case, the pobabltes of VM and bandwdth demand ae denoted by p(d) and p(b), beng ealzed n the second stage. The ealzaton nstance (b, d) eplaces the scenao ω that was used n the stochastc fomulaton. To ensue that the uncetan demand s accounted fo, the detemnstc fomulaton consdes each possble nstance of demand and chooses utlzaton and on-demand values fo each one, wth the total cost weghted by the pobablty of each ealzaton. We can also substtute n the new decson vaables and obtan new constants. Space does not pemt lstng each constant n full, but we gve an example fo llustaton. (b, d)+ (b, d) t, R,b B,d D (22) (22) s the detemnstc equvalent of (6). ω has been eplaced by the ealzaton of VM and bandwdth demand,.e., (b, d), and the equaton s evaluated fo each demand ealzaton, gven by b B and d D. The detemnstc equvalents of the emanng constants, (9)-(20), can be obtaned smlaly. The detemnstc equvalent fomulaton s a mxed ntege lnea pogammng poblem. Theefoe, a standad solve (e.g., CPLEX) can fnd an optmal soluton. V. PERFORMANCE EVALUATION A. Paamete Settng Ou fomulaton focuses on cost mnmzaton, wth a elatvely small envonment used fo pefomance evaluaton to demonstate optmalty. Resouce allocaton and ntege pogammng poblems ae NP-had, causng scalablty poblems n a cloud envonment wth a lage numbe of outes and VMs. The queston of computatonal effcency s an mpotant one to addess, pehaps though a dstbuted method, o a heustc elaxaton of the ntege constants. Investgaton of computatonal pefomance s left fo futue wok. 1) VMs and Cloud Povdes: We consde a epesentatve cloud computng envonment wth a sngle cloud use, fou cloud povdes, thee VM classes, and eght outes, aanged n two tes, as depcted n Fg. 2. Smla to [4], the demand equed fo each VM class anges fom 1 to 50,.e., D = {1, 2,...,50}. Fo smplcty, the sets of possble demands ae dentcal fo the thee VM classes as follows: D = {d 1,d 2,d 3 d 1 D 1,d 2 D 2,d 3 D 3,d 1 = d 2 = d 3 }. (23) The esouce equements fo each VM class ae dawn fom a synthess of easonable bandwdth equements, and the computng and stoage equements of the small, medum and lage standad nstances avalable fom Amazon EC2 [2]. Requements ae gven n daly quanttes, wth computng powe measued n CPU-hous, and stoage and bandwdth measued n GBs. Computng equements fo V 1, V 2, and V 3 ae 24, 24, and 48 CPU-hous, espectvely. Stoage equements fo classes V 1, V 2, and V 3 ae 160, 410, and 840, espectvely. Bandwdth equements fo V 1, V 2, and V 3 ae 3GB, 3GB, and 9GB pe day, espectvely. Compehensve pcng of esouces ae gven n Table I, whee the aconyms R, U, and O stand fo esevaton, utlzaton, and ondemand, espectvely. We hee explan the atonale fo these paamete settngs. As n [4], P 1 s consdeed as a pvate cloud mantaned by the use s ogansaton. The pvate cloud conssts of 10 seves, wth esevaton costs equal to the enegy costs of unnng the seves, gven by $357 pe seve pe yea. We take ths as the esevaton cost fo 24 CPUhous n the cloud. P 2 s pcng s the same as Amazon EC2, whlst P 3 s based on Wndows Azue s pcng schemes [11]. Wndows Azue does not have sepaate esevaton and utlzaton pcng, so we povde easonable values fo each. P 4 also follows [4] by only offeng an On-Demand plan, albet at cheape ates than the othe cloud povdes. The capacty of povdes P 2 to P 4 s taken to be unlmted, as they ae lage scale commecal cloud povdes. The computng capacty of P 1 s 240 CPU-hous, whch s the avalable computaton n a day fom 10 seves. The Stoage capacty s gven as 5000GB, based on the easonable assumpton that each seve has a 500GB had dsk dve. Intenal bandwdth fo the cloud 2972

5 IEEE ICC Next-Geneaton Netwokng Symposum 5 mn R c (e) (e) + p P c (e) Y (e) + R b B d D p(b)p(d)(c p P b B d D (b, d)+c p(b)p(d)(c (b, d))+ Y (b, d)+c Y (b, d)). (21) TABLE I VM COSTS FOR EACH CLOUD PROVIDER AND PROVISIONING PHASE. PRICES ARE GIVEN FOR A DAY S USAGE. Povde VM and Phase V 1 V 2 V 3 R U O R U O R U O P NA NA NA NA NA NA P P P 4 NA NA NA NA NA NA 7.68 TABLE II COST PER GB PER DAY OF BANDWIDTH PROVISIONING THROUGH ISP ROUTERS Routes Povsonng Phase R U O R 1,R R 2, R R 3, R 4, R 7, R Total Cost ($) Total cost 1st stage cost 2nd stage cost povdes s taken to be unlmted, as oute bandwdth lmts wll be stcte. 2) ISP Routes: Ou envonment has eght outes, wth the outes dvded nto thee sets of pcng, gven n Table II. Snce bandwdth esevaton and on demand pcng schemes ae not yet commonly avalable fom ISPs, we synthesze easonable values. The bandwdth pcng s based on the Sngtel 1 busness fbe boadband schemes [12]. Bandwdth pcng s gven pe GB pe day, wth the esevaton pcng fo each scheme adapted fom the monthly Gbps pces of the thee dffeent subsctons avalable fom Sngtel. Utlzaton pcng s a low cost acoss all outes and s chosen to be sutably popotonate to the esevaton pces. On-demand pces ae detemned as easonable values usng Sngtel s pe- GB excess usage fees fo moble boadband [13]. 3) Demand Uncetanty: We model the VM demand uncetanty as a Nomal dstbuton, wth μ = 76.5 and σ =36, wth bandwdth demand calculated as the sum of the bandwdth equements fo each equed VM. Fo smplcty, bandwdth demand s teated as a detemnstc functon of VM demand, so p(d) and p(b) ae equal, meanng that t s suffcent to evaluate only acoss one demand set. B. Numecal studes The detemnstc equvalent poblem above s encoded usng GAMS (Geneal Algebac Modelng System) [14]. 1) Impact of esevaton: We conduct two studes of the effects of esevaton n the system. In Fg. 3, we vay the numbe of eseved VMs and show the effect on total cost ncued to the use, both oveall, and n fst and second stages. 1 Sngtel s one of the lagest netwok opeatos n Sngapoe Numbe of eseved vtual machnes (VMs) Fg. 3. Cost of use when the numbe of eseved VMs s vaed. As expected, as the numbe of eseved VMs nceases, the fst stage cost nceases, and the second stage cost deceases. Ths esults n a dop of total pce untl the optmum pont s eached, when the benefts fom esevaton ae exceeded by the exta cost of esevng edundant VMs. We smlaly vay the amount of eseved bandwdth, but omt the gaph due to space constants. As expected, the bandwdth esevaton esult follows a smla patten to the VM esevaton esult. 2) Meetng Demand: To obseve ou soluton s effectveness n handlng vayng demands, we examne the costs unde dffeent demand ealzatons. We consde the total cost, the VM cost, and the bandwdth cost, and compae ou stochastc pogammng (SP) soluton s esults to a detemnstc veson (that knows the demand n advance), and an on-demand soluton, that has no esevaton opton. Fg. 4 shows the total costs of each soluton. As expected, the detemnstc soluton pefomed best, but ou soluton acheved pefomance that was close to the detemnstc soluton. The on-demand-only opton s sgnfcantly nfeo n all thee cases, although when the numbe of equed VMs s vey small t offes a lowe cost, because thee s no ovepovsonng cost. 3) Compason wth altenatve methods: In Fg. 5, we vay the mean of the pobablty dstbuton and compae ou stochastc pogammng (SP) soluton to two othe methods - one that employs no esevaton phase, and an Expected Value Fomulaton (EVF) that uses the mean as ts VM esevaton value. We also set the standad devaton of the demand dstbuton to 120 to bng out the dffeences n 2973

6 IEEE ICC Next-Geneaton Netwokng Symposum 6 Total cost ($) Stochastc pogammng Detemnstc On demand only Total cost ($) Povde P1 Povde P2 Povde P3 Povde P Numbe of equed vtual machnes (VMs) Fg. 4. Oveall costs, when the demand s known Cost of esevaton and on demand phase fom outes R7 & R8 ($) Fg. 6. Effect of oute pce vaaton on vtual machne (VM) placement 700 Total cost ($) Expected value functon (EVF) Stochastc pogammng On demand only Mean of vtual machne (VM) demand Fg. 5. Compang total cost of stochastc pogammng soluton wth altenatves pefomance between the methods. EVF pefoms well, as the optmal esevaton decson s close to the mean when usng nomally dstbuted demand. We antcate that f un aganst eal demand data, EVF would pefom sgnfcantly wose. As expected, on-demand-only s consdeably moe expensve. 4) Effect of vayng oute pces: To see the mpact of oute pcng on VM placement, we ase the pces fo outes R7 and R8. We set the esevaton and on-demand costs to be equal, as the dffeence n pce between VMs and bandwdth wll absob the exta cost of on-demand povsonng. In ths way, we can see that nceasng the pce of bandwdth povsonng foces mgaton of VMs fom P 3 and P 4 to altenatve cloud povdes. P 1 emans constant as t offes only esevaton. P 3 changes fst, wth VMs movng to P 2, whlst P 4 ntally stays nealy constant. Snce P 4 s ondemand only and offes the cheapest on-demand pces, t stll makes sense to use t. Howeve, once thee ae no moe VMs to emove fom P 3, the nceasng bandwdth costs affect P 4 as well, and the load s mgated to P 3, whee some of the taffc can be outed though the cheape R6. We show ths n Fg. 6 and obseve that even wth elatvely low pe-gb bandwdth pces, VM mgaton can be foced. VI. CONCLUSION We poposed an optmal algothm fo povsonng VMs and netwok bandwdth n a cloud computng envonment, enabled by softwae defned netwokng. We have accounted fo uncetan demand by fomulatng and solvng a two-stage stochastc optmzaton poblem to optmally eseve VMs and bandwdth to mnmze cost. We have evaluated ou soluton s pefomance n the descbed envonment. We have obseved that the soluton s optmal and outpefoms altenatve methods, achevng esults close to those acheved when demand s pefectly known. Futue dectons fo eseach nclude consdeng ISP netwoks that cay othe taffc and ntoduce a andom delay facto, consdeaton of non-detemnstc pevtual machne bandwdth demand, numecal studes based on eal usage data, and desgn of a dstbuted soluton fo lage-scale applcatons to mpove scalablty. REFERENCES [1] Softwae-defned netwokng: The new nom fo netwoks, Whte Pape, Apl [Onlne]. Avalable: sdn-esouces/whte-papes/wp-sdn-newnom.pdf [2] (2013, Aug.) Amazon ec2. [Onlne]. Avalable: ec2/pcng/ [3] L. Gt, D. Iwn, A. Yumeefend, and J. Chase, Vtual machne hostng fo netwoked clustes: Buldng the foundatons fo autonomc ochestaton, n Vtualzaton Technology n Dstbuted Computng, VTDC Fst Intenatonal Wokshop on, 2006, pp [4] S. Chas, B.-S. Lee, and D. Nyato, Optmzaton of esouce povsonng cost n cloud computng, Sevces Computng, IEEE Tansactons on, vol. 5, no. 2, pp , [5] J. R. Bge and F. Louveaux, Intoducton to Stochastc Pogammng. New Yok: Spnge, [6] Q. T. Nguyen, N. Quang-Hung, N. H. Tuong, V. H. Tan, and N. Thoa, Vtual machne allocaton n cloud computng fo mnmzng total executon tme on each machne, n Computng, Management and Telecommuncatons (ComManTel), 2013 Intenatonal Confeence on, Jan 2013, pp [7] H. Xu and B. L, Jont equest mappng and esponse outng fo geo-dstbuted cloud sevces, n INFOCOM, 2013 Poceedngs IEEE, 2013, pp [8] N. McKeown, T. Andeson, H. Balakshnan, G. Paulka, L. Peteson, J. Rexfod, S. Shenke, and J. Tune, Openflow: Enablng nnovaton n campus netwoks, SIGCOMM Comput. Commun. Rev., vol. 38, no. 2, pp , Ma [Onlne]. Avalable: [9] M. Chowdhuy, M. Rahman, and R. Boutaba, Vneyad: Vtual netwok embeddng algothms wth coodnated node and lnk mappng, Netwokng, IEEE/ACM Tansactons on, vol. 20, no. 1, pp , Feb [10] H. D, L. L, V. Anand, H. Yu, and G. Sun, Cost effcent vtual nfastuctue mappng usng subgaph somophsm, n Communcatons and Photoncs Confeence and Exhbton (ACP), 2010 Asa, Dec 2010, pp [11] Wndows azue, Aug [Onlne]. Avalable: wndowsazue.com/en-us/pcng/detals/vtual-machnes/ [12] (2013, Aug.) Sngtel busness. [Onlne]. Avalable: sngnet-evolve-fbe-boadband [13] (2013, Sep.) Sngtel moble boadband faqs. [Onlne]. Avalable: ntenet/boadband-on-the-go/moble-boadband/moble/faq [14] D. Chattopadhyay, Applcaton of geneal algebac modelng system to powe system optmzaton, Powe Systems, IEEE Tansactons on, vol. 14, no. 1, pp ,

A Resource Scheduling Algorithms Based on the Minimum Relative Degree of Load Imbalance

A Resource Scheduling Algorithms Based on the Minimum Relative Degree of Load Imbalance Jounal of Communcatons Vol. 10, No. 10, Octobe 2015 A Resouce Schedulng Algothms Based on the Mnmum Relatve Degee of Load Imbalance Tao Xue and Zhe Fan Depatment of Compute Scence, X an Polytechnc Unvesty,

More information

Additional File 1 - A model-based circular binary segmentation algorithm for the analysis of array CGH data

Additional File 1 - A model-based circular binary segmentation algorithm for the analysis of array CGH data 1 Addtonal Fle 1 - A model-based ccula bnay segmentaton algothm fo the analyss of aay CGH data Fang-Han Hsu 1, Hung-I H Chen, Mong-Hsun Tsa, Lang-Chuan La 5, Ch-Cheng Huang 1,6, Shh-Hsn Tu 6, Ec Y Chuang*

More information

A New replenishment Policy in a Two-echelon Inventory System with Stochastic Demand

A New replenishment Policy in a Two-echelon Inventory System with Stochastic Demand A ew eplenshment Polcy n a wo-echelon Inventoy System wth Stochastc Demand Rasoul Haj, Mohammadal Payesh eghab 2, Amand Babol 3,2 Industal Engneeng Dept, Shaf Unvesty of echnology, ehan, Ian (haj@shaf.edu,

More information

AREA COVERAGE SIMULATIONS FOR MILLIMETER POINT-TO-MULTIPOINT SYSTEMS USING STATISTICAL MODEL OF BUILDING BLOCKAGE

AREA COVERAGE SIMULATIONS FOR MILLIMETER POINT-TO-MULTIPOINT SYSTEMS USING STATISTICAL MODEL OF BUILDING BLOCKAGE Radoengneeng Aea Coveage Smulatons fo Mllmete Pont-to-Multpont Systems Usng Buldng Blockage 43 Vol. 11, No. 4, Decembe AREA COVERAGE SIMULATIONS FOR MILLIMETER POINT-TO-MULTIPOINT SYSTEMS USING STATISTICAL

More information

Keywords: Transportation network, Hazardous materials, Risk index, Routing, Network optimization.

Keywords: Transportation network, Hazardous materials, Risk index, Routing, Network optimization. IUST Intenatonal Jounal of Engneeng Scence, Vol. 19, No.3, 2008, Page 57-65 Chemcal & Cvl Engneeng, Specal Issue A ROUTING METHODOLOGY FOR HAARDOUS MATIALS TRANSPORTATION TO REDUCE THE RISK OF ROAD NETWORK

More information

A Novel Lightweight Algorithm for Secure Network Coding

A Novel Lightweight Algorithm for Secure Network Coding A Novel Lghtweght Algothm fo Secue Netwok Codng A Novel Lghtweght Algothm fo Secue Netwok Codng State Key Laboatoy of Integated Sevce Netwoks, Xdan Unvesty, X an, Chna, E-mal: {wangxaoxao,wangmeguo}@mal.xdan.edu.cn

More information

Perturbation Theory and Celestial Mechanics

Perturbation Theory and Celestial Mechanics Copyght 004 9 Petubaton Theoy and Celestal Mechancs In ths last chapte we shall sketch some aspects of petubaton theoy and descbe a few of ts applcatons to celestal mechancs. Petubaton theoy s a vey boad

More information

An Algorithm For Factoring Integers

An Algorithm For Factoring Integers An Algothm Fo Factong Integes Yngpu Deng and Yanbn Pan Key Laboatoy of Mathematcs Mechanzaton, Academy of Mathematcs and Systems Scence, Chnese Academy of Scences, Bejng 100190, People s Republc of Chna

More information

Efficient Evolutionary Data Mining Algorithms Applied to the Insurance Fraud Prediction

Efficient Evolutionary Data Mining Algorithms Applied to the Insurance Fraud Prediction Intenatonal Jounal of Machne Leanng and Computng, Vol. 2, No. 3, June 202 Effcent Evolutonay Data Mnng Algothms Appled to the Insuance Faud Pedcton Jenn-Long Lu, Chen-Lang Chen, and Hsng-Hu Yang Abstact

More information

Mixed Task Scheduling and Resource Allocation Problems

Mixed Task Scheduling and Resource Allocation Problems Task schedulng and esouce allocaton 1 Mxed Task Schedulng and Resouce Allocaton Poblems Mae-José Huguet 1,2 and Pee Lopez 1 1 LAAS-CNRS, 7 av. du Colonel Roche F-31077 Toulouse cedex 4, Fance {huguet,lopez}@laas.f

More information

Gravitation. Definition of Weight Revisited. Newton s Law of Universal Gravitation. Newton s Law of Universal Gravitation. Gravitational Field

Gravitation. Definition of Weight Revisited. Newton s Law of Universal Gravitation. Newton s Law of Universal Gravitation. Gravitational Field Defnton of Weght evsted Gavtaton The weght of an object on o above the eath s the gavtatonal foce that the eath exets on the object. The weght always ponts towad the cente of mass of the eath. On o above

More information

Bending Stresses for Simple Shapes

Bending Stresses for Simple Shapes -6 Bendng Stesses fo Smple Sapes In bendng, te maxmum stess and amount of deflecton can be calculated n eac of te followng stuatons. Addtonal examples ae avalable n an engneeng andbook. Secton Modulus

More information

On the Efficiency of Equilibria in Generalized Second Price Auctions

On the Efficiency of Equilibria in Generalized Second Price Auctions On the Effcency of Equlba n Genealzed Second Pce Auctons Ioanns Caaganns Panagots Kanellopoulos Chstos Kaklamans Maa Kyopoulou Depatment of Compute Engneeng and Infomatcs Unvesty of Patas and RACTI, Geece

More information

TRUCK ROUTE PLANNING IN NON- STATIONARY STOCHASTIC NETWORKS WITH TIME-WINDOWS AT CUSTOMER LOCATIONS

TRUCK ROUTE PLANNING IN NON- STATIONARY STOCHASTIC NETWORKS WITH TIME-WINDOWS AT CUSTOMER LOCATIONS TRUCK ROUTE PLANNING IN NON- STATIONARY STOCHASTIC NETWORKS WITH TIME-WINDOWS AT CUSTOMER LOCATIONS Hossen Jula α, Maged Dessouky β, and Petos Ioannou γ α School of Scence, Engneeng and Technology, Pennsylvana

More information

Electric Potential. otherwise to move the object from initial point i to final point f

Electric Potential. otherwise to move the object from initial point i to final point f PHY2061 Enched Physcs 2 Lectue Notes Electc Potental Electc Potental Dsclame: These lectue notes ae not meant to eplace the couse textbook. The content may be ncomplete. Some topcs may be unclea. These

More information

PREVENTIVE AND CORRECTIVE SECURITY MARKET MODEL

PREVENTIVE AND CORRECTIVE SECURITY MARKET MODEL REVENTIVE AND CORRECTIVE SECURITY MARKET MODEL Al Ahmad-hat Rachd Cheaou and Omd Alzadeh Mousav Ecole olytechnque Fédéale de Lausanne Lausanne Swzeland al.hat@epfl.ch achd.cheaou@epfl.ch omd.alzadeh@epfl.ch

More information

Charging the Internet Without Bandwidth Reservation: An Overview and Bibliography of Mathematical Approaches

Charging the Internet Without Bandwidth Reservation: An Overview and Bibliography of Mathematical Approaches JOURNAL OF INFORMATION SCIENCE AND ENGINEERING 9, -xxx (2003) Chagng the Intenet Wthout Bandwdth Resevaton: An Ovevew and Bblogaphy of Mathematcal Appoaches IRISA-INRIA Campus Unvestae de Beauleu 35042

More information

Orbit dynamics and kinematics with full quaternions

Orbit dynamics and kinematics with full quaternions bt dynamcs and knematcs wth full quatenons Davde Andes and Enco S. Canuto, Membe, IEEE Abstact Full quatenons consttute a compact notaton fo descbng the genec moton of a body n the space. ne of the most

More information

Prejudice and the Economics of Discrimination

Prejudice and the Economics of Discrimination Pelmnay Pejudce and the Economcs of Dscmnaton Kewn Kof Chales Unvesty of Chcago and NB Jonathan Guyan Unvesty of Chcago GSB and NB Novembe 17, 2006 Abstact Ths pape e-examnes the ole of employe pejudce

More information

econstor zbw www.econstor.eu

econstor zbw www.econstor.eu econsto www.econsto.eu De Open-Access-Publkatonsseve de ZBW Lebnz-Infomatonszentum Wtschaft The Open Access Publcaton Seve of the ZBW Lebnz Infomaton Cente fo Economcs Babazadeh, Reza; Razm, Jafa; Ghods,

More information

Fault tolerance in cloud technologies presented as a service

Fault tolerance in cloud technologies presented as a service Internatonal Scentfc Conference Computer Scence 2015 Pavel Dzhunev, PhD student Fault tolerance n cloud technologes presented as a servce INTRODUCTION Improvements n technques for vrtualzaton and performance

More information

PCA vs. Varimax rotation

PCA vs. Varimax rotation PCA vs. Vamax otaton The goal of the otaton/tansfomaton n PCA s to maxmze the vaance of the new SNP (egensnp), whle mnmzng the vaance aound the egensnp. Theefoe the dffeence between the vaances captued

More information

A Mathematical Model for Selecting Third-Party Reverse Logistics Providers

A Mathematical Model for Selecting Third-Party Reverse Logistics Providers A Mathematcal Model fo Selectng Thd-Pat Revese Logstcs Povdes Reza Fazpoo Saen Depatment of Industal Management, Facult of Management and Accountng, Islamc Azad Unvest - Kaaj Banch, Kaaj, Ian, P. O. Box:

More information

(Semi)Parametric Models vs Nonparametric Models

(Semi)Parametric Models vs Nonparametric Models buay, 2003 Pobablty Models (Sem)Paametc Models vs Nonpaametc Models I defne paametc, sempaametc, and nonpaametc models n the two sample settng My defnton of sempaametc models s a lttle stonge than some

More information

Research on Cloud Computing Load Balancing Based on Virtual Machine Migration

Research on Cloud Computing Load Balancing Based on Virtual Machine Migration Send Odes fo Repnts to epnts@benthascence.ae 334 The Open Cybenetcs & Systecs Jounal, 205, 9, 334-340 Open Access Reseach on Cloud Coputng Load Balancng Based on Vtual Machne Mgaton Lu Kun,*, Xu Gaochao

More information

A Coverage Gap Filling Algorithm in Hybrid Sensor Network

A Coverage Gap Filling Algorithm in Hybrid Sensor Network A Coveage Ga Fllng Algothm n Hybd Senso Netwok Tan L, Yang Mnghua, Yu Chongchong, L Xuanya, Cheng Bn A Coveage Ga Fllng Algothm n Hybd Senso Netwok 1 Tan L, 2 Yang Mnghua, 3 Yu Chongchong, 4 L Xuanya,

More information

Competitive Targeted Advertising with Price Discrimination

Competitive Targeted Advertising with Price Discrimination Compette Tageted Adetsng wth Pce Dscmnaton Rosa Banca Estees Unesdade do Mnho and NIPE banca@eeg.umnho.pt Joana Resende Faculdade de Economa, Unesdade do Poto and CEF.UP jesende@fep.up.pt Septembe 8, 205

More information

AN EQUILIBRIUM ANALYSIS OF THE INSURANCE MARKET WITH VERTICAL DIFFERENTIATION

AN EQUILIBRIUM ANALYSIS OF THE INSURANCE MARKET WITH VERTICAL DIFFERENTIATION QUIIRIUM YI OF T IUR MRKT WIT VRTI IFFRTITIO Mahto Okua Faculty of conomcs, agasak Unvesty, 4-- Katafuch, agasak, 8508506, Japan okua@net.nagasak-u.ac.p TRT ach nsuance poduct pe se s dentcal but the nsuance

More information

Statistical modelling of gambling probabilities

Statistical modelling of gambling probabilities Ttle Statstcal modellng of gamblng pobabltes Autho(s) Lo, Su-yan, Vcto.; 老 瑞 欣 Ctaton Issued Date 992 URL http://hdl.handle.net/0722/3525 Rghts The autho etans all popetay ghts, (such as patent ghts) and

More information

RESOURCE SCHEDULING STRATEGIES FOR A NETWORK-BASED AUTONOMOUS MOBILE ROBOT. Hongryeol Kim, Joomin Kim, and Daewon Kim

RESOURCE SCHEDULING STRATEGIES FOR A NETWORK-BASED AUTONOMOUS MOBILE ROBOT. Hongryeol Kim, Joomin Kim, and Daewon Kim RESOURCE SCHEDULING SRAEGIES FOR A NEWORK-BASED AUONOMOUS MOBILE ROBO Hongyeol K, Joon K, and Daewon K Depatent of Infoaton Contol Engneeng, Myong Unvesty Abstact: In ths pape, effcent esouce schedulng

More information

REAL TIME MONITORING OF DISTRIBUTION NETWORKS USING INTERNET BASED PMU. Akanksha Eknath Pachpinde

REAL TIME MONITORING OF DISTRIBUTION NETWORKS USING INTERNET BASED PMU. Akanksha Eknath Pachpinde REAL TME MONTORNG OF DSTRBUTON NETWORKS USNG NTERNET BASED PMU by Akanksha Eknath Pachpnde A Thess submtted to the Faculty of the Gaduate School of the Unvesty at Buffalo, State Unvesty of New Yok n patal

More information

The Can-Order Policy for One-Warehouse N-Retailer Inventory System: A Heuristic Approach

The Can-Order Policy for One-Warehouse N-Retailer Inventory System: A Heuristic Approach Atcle Te Can-Ode Polcy fo One-Waeouse N-Retale Inventoy ystem: A Heustc Appoac Vaapon Pukcanon, Paveena Caovaltongse, and Naagan Pumcus Depatment of Industal Engneeng, Faculty of Engneeng, Culalongkon

More information

UNIVERSIDAD DE LA REPÚBLICA Facultad de Ingeniería RESOURCE ALLOCATION IN NETWORKS FROM A CONNECTION-LEVEL PERSPECTIVE

UNIVERSIDAD DE LA REPÚBLICA Facultad de Ingeniería RESOURCE ALLOCATION IN NETWORKS FROM A CONNECTION-LEVEL PERSPECTIVE UNIVERSIDAD DE LA REPÚBLICA Facultad de Ingeneía Tess paa opta al Título de Docto en Ingeneía Eléctca RESOURCE ALLOCATION IN NETWORKS FROM A CONNECTION-LEVEL PERSPECTIVE (ASIGNACIÓN DE RECURSOS EN REDES

More information

Statistical Discrimination or Prejudice? A Large Sample Field Experiment. Michael Ewens, Bryan Tomlin, and Liang Choon Wang.

Statistical Discrimination or Prejudice? A Large Sample Field Experiment. Michael Ewens, Bryan Tomlin, and Liang Choon Wang. Statstcal Dscmnaton o Pejudce? A Lage Sample Feld Expement Mchael Ewens, yan Tomln, and Lang Choon ang Abstact A model of acal dscmnaton povdes testable mplcatons fo two featues of statstcal dscmnatos:

More information

Institut für Halle Institute for Economic Research Wirtschaftsforschung Halle

Institut für Halle Institute for Economic Research Wirtschaftsforschung Halle Insttut fü Halle Insttute fo Economc Reseach Wtschaftsfoschung Halle A Smple Repesentaton of the Bea-Jaque-Lee Test fo Pobt Models Joachm Wlde Dezembe 2007 No. 13 IWH-Dskussonspapee IWH-Dscusson Papes

More information

Discussion Papers. Thure Traber Claudia Kemfert

Discussion Papers. Thure Traber Claudia Kemfert Dscusson Papes Thue Tabe Clauda Kemfet Impacts of the Geman Suppot fo Renewable Enegy on Electcty Pces, Emssons and Pofts: An Analyss Based on a Euopean Electcty Maket Model Beln, July 2007 Opnons expessed

More information

LINES ON BRIESKORN-PHAM SURFACES

LINES ON BRIESKORN-PHAM SURFACES LIN ON BRIKORN-PHAM URFAC GUANGFNG JIANG, MUTUO OKA, DUC TAI PHO, AND DIRK IRMA Abstact By usng toc modfcatons and a esult of Gonzalez-pnbeg and Lejeune- Jalabet, we answe the followng questons completely

More information

The Greedy Method. Introduction. 0/1 Knapsack Problem

The Greedy Method. Introduction. 0/1 Knapsack Problem The Greedy Method Introducton We have completed data structures. We now are gong to look at algorthm desgn methods. Often we are lookng at optmzaton problems whose performance s exponental. For an optmzaton

More information

FAIR VALUATION OF VARIOUS PARTICIPATION SCHEMES IN LIFE INSURANCE ABSTRACT

FAIR VALUATION OF VARIOUS PARTICIPATION SCHEMES IN LIFE INSURANCE ABSTRACT FAIR VALUAION OF VARIOUS PARIIPAION SHEMES IN LIFE INSURANE PIERRE DEVOLDER AND INMAULADA DOMÍNGUEZ-FABIÁN BY Insttut des Scences Actuaelles, Unvesté atholque de Louvan, 6 ue des Wallons, 348 Louvan la

More information

CONSTRUCTION PROJECT SCHEDULING WITH IMPRECISELY DEFINED CONSTRAINTS

CONSTRUCTION PROJECT SCHEDULING WITH IMPRECISELY DEFINED CONSTRAINTS Management an Innovaton fo a Sustanable Bult Envonment ISBN: 9789052693958 20 23 June 2011, Amsteam, The Nethelans CONSTRUCTION PROJECT SCHEDULING WITH IMPRECISELY DEFINED CONSTRAINTS JANUSZ KULEJEWSKI

More information

Statistical Discrimination or Prejudice? A Large Sample Field Experiment

Statistical Discrimination or Prejudice? A Large Sample Field Experiment DEPARTMENT OF ECONOMICS ISSN 1441-5429 DISCUSSION PAPER 23/12 Statstcal Dscmnaton o Pejudce? A Lage Sample Feld Expement Mchael Ewens, yan Tomln, and Lang Choon ang * Abstact A model of acal dscmnaton

More information

Determinants of Borrowing Limits on Credit Cards Shubhasis Dey and Gene Mumy

Determinants of Borrowing Limits on Credit Cards Shubhasis Dey and Gene Mumy Bank of Canada Banque du Canada Wokng Pape 2005-7 / Document de taval 2005-7 Detemnants of Boowng mts on Cedt Cads by Shubhass Dey and Gene Mumy ISSN 1192-5434 Pnted n Canada on ecycled pape Bank of Canada

More information

Security of Full-State Keyed Sponge and Duplex: Applications to Authenticated Encryption

Security of Full-State Keyed Sponge and Duplex: Applications to Authenticated Encryption Secuty of Full-State Keyed Sponge and uplex: Applcatons to Authentcated Encypton Bat Mennnk 1 Reza Reyhantaba 2 aman Vzá 2 1 ept. Electcal Engneeng, ESAT/COSIC, KU Leuven, and Mnds, Belgum bat.mennnk@esat.kuleuven.be

More information

Survey on Virtual Machine Placement Techniques in Cloud Computing Environment

Survey on Virtual Machine Placement Techniques in Cloud Computing Environment Survey on Vrtual Machne Placement Technques n Cloud Computng Envronment Rajeev Kumar Gupta and R. K. Paterya Department of Computer Scence & Engneerng, MANIT, Bhopal, Inda ABSTRACT In tradtonal data center

More information

A New Estimation Model for Small Organic Software Project

A New Estimation Model for Small Organic Software Project 8 03 ACADEMY PUBLISHER A New Estmaton Model o Small Oganc Sotwae Poject Wan-Jang Han, Tan-Bo Lu, and Xao-Yan Zhang School O Sotwae Engneeng, Bejng Unvesty o Posts and Telecommuncaton, Bejng, Chna Emal:

More information

Comparing Availability of Various Rack Power Redundancy Configurations

Comparing Availability of Various Rack Power Redundancy Configurations Compaing Availability of Vaious Rack Powe Redundancy Configuations By Victo Avela White Pape #48 Executive Summay Tansfe switches and dual-path powe distibution to IT equipment ae used to enhance the availability

More information

Drag force acting on a bubble in a cloud of compressible spherical bubbles at large Reynolds numbers

Drag force acting on a bubble in a cloud of compressible spherical bubbles at large Reynolds numbers Euopean Jounal of Mechancs B/Fluds 24 2005 468 477 Dag foce actng on a bubble n a cloud of compessble sphecal bubbles at lage Reynolds numbes S.L. Gavlyuk a,b,,v.m.teshukov c a Laboatoe de Modélsaton en

More information

Molecular Dynamics. r F. r dt. What is molecular dynamics?

Molecular Dynamics. r F. r dt. What is molecular dynamics? What s molecula dynamcs? Molecula Dynamcs Molecula dynamcs (MD) s a compute smulaton technque that allows one to pedct the tme evoluton of a system of nteactng patcles (atoms, molecules, ganules, etc.).

More information

International Business Cycles and Exchange Rates

International Business Cycles and Exchange Rates Revew of Intenatonal Economcs, 7(4), 682 698, 999 Intenatonal Busness Cycles and Exchange Rates Chstan Zmmemann* Abstact Models of ntenatonal eal busness cycles ae not able to account fo the hgh volatlty

More information

A PARTICLE-BASED LAGRANGIAN CFD TOOL FOR FREE-SURFACE SIMULATION

A PARTICLE-BASED LAGRANGIAN CFD TOOL FOR FREE-SURFACE SIMULATION C A N A L D E E X P E R I E N C I A S H I D R O D I N Á M I C A S, E L P A R D O Publcacón núm. 194 A PARTICLE-BASED LAGRANGIAN CFD TOOL FOR FREE-SURFACE SIMULATION POR D. MUÑOZ V. GONZÁLEZ M. BLAIN J.

More information

The Costs of Disposal and Recycling. An Application to Italian Municipal Solid Waste Services

The Costs of Disposal and Recycling. An Application to Italian Municipal Solid Waste Services The Costs of Dsposal and Recyclng. An Applcaton to Italan Muncpal Sold Waste Sevces Gazano Abate Fabzo Ebetta Govann Faquell Davde Vannon No. 3 Decembe 011 www.caloalbeto.og/wokng_papes 011 by Gazano Abate,

More information

Evaluating the impact of Blade Server and Virtualization Software Technologies on the RIT Datacenter

Evaluating the impact of Blade Server and Virtualization Software Technologies on the RIT Datacenter Evaluating the impact of and Vitualization Softwae Technologies on the RIT Datacente Chistophe M Butle Vitual Infastuctue Administato Rocheste Institute of Technology s Datacente Contact: chis.butle@it.edu

More information

Comparing Availability of Various Rack Power Redundancy Configurations

Comparing Availability of Various Rack Power Redundancy Configurations Compaing Availability of Vaious Rack Powe Redundancy Configuations White Pape 48 Revision by Victo Avela > Executive summay Tansfe switches and dual-path powe distibution to IT equipment ae used to enhance

More information

Green's function integral equation methods for plasmonic nanostructures

Green's function integral equation methods for plasmonic nanostructures Geens functon ntegal equaton methods fo plasmonc nanostuctues (Ph Couse: Optcal at the Nanoscale) Thomas Søndegaad epatment of Phscs and Nanotechnolog, Aalbog Unvest, Senve 4A, K-9 Aalbog Øst, enma. Intoducton

More information

Design and Implementation of a Smart LED Lighting System Using a Self Adaptive Weighted Data Fusion Algorithm

Design and Implementation of a Smart LED Lighting System Using a Self Adaptive Weighted Data Fusion Algorithm Sensos 2013, 13, 16915-16939; do:10.3390/s131216915 Atcle OPEN ACCESS sensos ISSN 1424-8220 www.mdp.com/jounal/sensos Desgn and Implementaton of a Smat LED Lghtng System Usng a Self Adaptve Weghted Data

More information

Effect of Contention Window on the Performance of IEEE 802.11 WLANs

Effect of Contention Window on the Performance of IEEE 802.11 WLANs Effect of Contention Window on the Pefomance of IEEE 82.11 WLANs Yunli Chen and Dhama P. Agawal Cente fo Distibuted and Mobile Computing, Depatment of ECECS Univesity of Cincinnati, OH 45221-3 {ychen,

More information

An Approach to Optimized Resource Allocation for Cloud Simulation Platform

An Approach to Optimized Resource Allocation for Cloud Simulation Platform An Appoach to Optimized Resouce Allocation fo Cloud Simulation Platfom Haitao Yuan 1, Jing Bi 2, Bo Hu Li 1,3, Xudong Chai 3 1 School of Automation Science and Electical Engineeing, Beihang Univesity,

More information

MOTION PLANNING FOR MULTI-ROBOT ASSEMBLY SYSTEMS. M. Bonert L.H. Shu B. Benhabib*

MOTION PLANNING FOR MULTI-ROBOT ASSEMBLY SYSTEMS. M. Bonert L.H. Shu B. Benhabib* Poceengs of the 999 ASME Desgn Engneeng Techncal Confeences Septembe -, 999, Las Vegas, Nevaa DETC99/DAC-869 MOTION PLANNING FOR MULTI-ROBOT ASSEMBLY SYSTEMS M. Bonet L.H. Shu B. Benhabb* Compute Integate

More information

est using the formula I = Prt, where I is the interest earned, P is the principal, r is the interest rate, and t is the time in years.

est using the formula I = Prt, where I is the interest earned, P is the principal, r is the interest rate, and t is the time in years. 9.2 Inteest Objectives 1. Undestand the simple inteest fomula. 2. Use the compound inteest fomula to find futue value. 3. Solve the compound inteest fomula fo diffeent unknowns, such as the pesent value,

More information

Questions & Answers Chapter 10 Software Reliability Prediction, Allocation and Demonstration Testing

Questions & Answers Chapter 10 Software Reliability Prediction, Allocation and Demonstration Testing M13914 Questions & Answes Chapte 10 Softwae Reliability Pediction, Allocation and Demonstation Testing 1. Homewok: How to deive the fomula of failue ate estimate. λ = χ α,+ t When the failue times follow

More information

Continuous Compounding and Annualization

Continuous Compounding and Annualization Continuous Compounding and Annualization Philip A. Viton Januay 11, 2006 Contents 1 Intoduction 1 2 Continuous Compounding 2 3 Pesent Value with Continuous Compounding 4 4 Annualization 5 5 A Special Poblem

More information

Tasmanian Greens MPs Reply to the Electricity Supply Industry Expert Panel: Powering Tasmania s Transition to a 100% Renewable Energy Future

Tasmanian Greens MPs Reply to the Electricity Supply Industry Expert Panel: Powering Tasmania s Transition to a 100% Renewable Energy Future T a s ma n a ng eensmps Repl yt ot hee l ec t c t ys uppl yi ndus t ye x pe tpa nel Powe ngat ans t ont oa1 0 0 %Re ne wabl ee ne gyf ut u e Rel ea s edaugus t2012 Aut ho s edbyn c kmc K m MP, Pa l a menthous

More information

4. SHAFT SENSORLESS FORCED DYNAMICS CONTROL OF RELUCTANCE SYNCHRONOUS MOTOR DRIVES

4. SHAFT SENSORLESS FORCED DYNAMICS CONTROL OF RELUCTANCE SYNCHRONOUS MOTOR DRIVES 4. SHAFT SENSORLESS FORCED DYNAMICS CONTROL OF RELUCTANCE SYNCHRONOUS MOTOR DRIVES 4.. VECTOR CONTROLLED RELUCTANCE SYNCHRONOUS MOTOR DRIVES WITH PRESCRIBED CLOSED-LOOP SPEED DYNAMICS Abstact: A new spee

More information

HEALTHCARE INTEGRATION BASED ON CLOUD COMPUTING

HEALTHCARE INTEGRATION BASED ON CLOUD COMPUTING U.P.B. Sci. Bull., Seies C, Vol. 77, Iss. 2, 2015 ISSN 2286-3540 HEALTHCARE INTEGRATION BASED ON CLOUD COMPUTING Roxana MARCU 1, Dan POPESCU 2, Iulian DANILĂ 3 A high numbe of infomation systems ae available

More information

How To Use A Network On A Network With A Powerline (Lan) On A Pcode (Lan On Alan) (Lan For Acedo) (Moe) (Omo) On An Ipo) Or Ipo (

How To Use A Network On A Network With A Powerline (Lan) On A Pcode (Lan On Alan) (Lan For Acedo) (Moe) (Omo) On An Ipo) Or Ipo ( Hubs, Bidges, and Switches Used fo extending LANs in tems of geogaphical coveage, numbe of nodes, administation capabilities, etc. Diffe in egads to: m collision domain isolation m laye at which they opeate

More information

I = Prt. = P(1+i) n. A = Pe rt

I = Prt. = P(1+i) n. A = Pe rt 11 Chapte 6 Matheatcs of Fnance We wll look at the atheatcs of fnance. 6.1 Sple and Copound Inteest We wll look at two ways nteest calculated on oney. If pncpal pesent value) aount P nvested at nteest

More information

Fair Virtual Bandwidth Allocation Model in Virtual Data Centers

Fair Virtual Bandwidth Allocation Model in Virtual Data Centers Far Vrtual Bandwdth Allocaton Model n Vrtual Data Centers Yng Yuan, Cu-rong Wang, Cong Wang School of Informaton Scence and Engneerng ortheastern Unversty Shenyang, Chna School of Computer and Communcaton

More information

Modeling and computing constrained

Modeling and computing constrained F EAURE A RICLE HE COMPUAION OF CONSRAINED DYNAMICAL SYSEMS: MACHING PHYSICAL MODELING WIH NUMERICAL MEHODS Reseaches have nvestgated modelng and computaton of constaned dynamcal systems, but scentsts

More information

How a Global Inter-Country Input-Output Table with Processing Trade Account. Can be constructed from GTAP Database

How a Global Inter-Country Input-Output Table with Processing Trade Account. Can be constructed from GTAP Database How a lobal Inte-County Input-Output Table wth Pocessng Tade Account Can be constucted fom TAP Database Manos Tsgas and Zh Wang U.S. Intenatonal Tade Commsson* Mak ehlha U.S. Depatment of Inteo* (Pelmnay

More information

Impact on inventory costs with consolidation of distribution centers

Impact on inventory costs with consolidation of distribution centers IIE Tansactons (2) 33, 99± Imact on nventoy costs wth consoldaton of dstbuton centes CHUNG PIAW TEO, JIHONG OU and MARK GOH Deatment of Decson Scences, Faculty of Busness Admnstaton, Natonal Unvesty of

More information

Efficient Redundancy Techniques for Latency Reduction in Cloud Systems

Efficient Redundancy Techniques for Latency Reduction in Cloud Systems Efficient Redundancy Techniques fo Latency Reduction in Cloud Systems 1 Gaui Joshi, Emina Soljanin, and Gegoy Wonell Abstact In cloud computing systems, assigning a task to multiple seves and waiting fo

More information

Department of Economics Working Paper Series

Department of Economics Working Paper Series Depatment of Economcs Wokng Pape Sees Reputaton and Effcency: A Nonpaametc Assessment of Ameca s Top-Rated MBA Pogams Subhash C. Ray Unvesty of Connectcut Yongl Jeon Cental Mchgan Unvest Wokng Pape 23-3

More information

How To Solve An Onlne Control Polcy On A Vrtualzed Data Center

How To Solve An Onlne Control Polcy On A Vrtualzed Data Center Dynamc Resource Allocaton and Power Management n Vrtualzed Data Centers Rahul Urgaonkar, Ulas C. Kozat, Ken Igarash, Mchael J. Neely urgaonka@usc.edu, {kozat, garash}@docomolabs-usa.com, mjneely@usc.edu

More information

Ilona V. Tregub, ScD., Professor

Ilona V. Tregub, ScD., Professor Investment Potfolio Fomation fo the Pension Fund of Russia Ilona V. egub, ScD., Pofesso Mathematical Modeling of Economic Pocesses Depatment he Financial Univesity unde the Govenment of the Russian Fedeation

More information

Network Aware Load-Balancing via Parallel VM Migration for Data Centers

Network Aware Load-Balancing via Parallel VM Migration for Data Centers Network Aware Load-Balancng va Parallel VM Mgraton for Data Centers Kun-Tng Chen 2, Chen Chen 12, Po-Hsang Wang 2 1 Informaton Technology Servce Center, 2 Department of Computer Scence Natonal Chao Tung

More information

An Alternative Way to Measure Private Equity Performance

An Alternative Way to Measure Private Equity Performance An Alternatve Way to Measure Prvate Equty Performance Peter Todd Parlux Investment Technology LLC Summary Internal Rate of Return (IRR) s probably the most common way to measure the performance of prvate

More information

Alarm transmission through Radio and GSM networks

Alarm transmission through Radio and GSM networks Alam tansmission though Radio and GSM netwoks 2015 Alam tansmission though Radio netwok RR-IP12 RL10 E10C E10C LAN RL1 0 R11 T10 (T10U) Windows MONAS MS NETWORK MCI > GNH > GND > +E > DATA POWER DATA BUS

More information

2 r2 θ = r2 t. (3.59) The equal area law is the statement that the term in parentheses,

2 r2 θ = r2 t. (3.59) The equal area law is the statement that the term in parentheses, 3.4. KEPLER S LAWS 145 3.4 Keple s laws You ae familia with the idea that one can solve some mechanics poblems using only consevation of enegy and (linea) momentum. Thus, some of what we see as objects

More information

Support Vector Machines

Support Vector Machines Support Vector Machnes Max Wellng Department of Computer Scence Unversty of Toronto 10 Kng s College Road Toronto, M5S 3G5 Canada wellng@cs.toronto.edu Abstract Ths s a note to explan support vector machnes.

More information

Chapter 3 Savings, Present Value and Ricardian Equivalence

Chapter 3 Savings, Present Value and Ricardian Equivalence Chapte 3 Savings, Pesent Value and Ricadian Equivalence Chapte Oveview In the pevious chapte we studied the decision of households to supply hous to the labo maket. This decision was a static decision,

More information

Institute of Informatics, Faculty of Business and Management, Brno University of Technology,Czech Republic

Institute of Informatics, Faculty of Business and Management, Brno University of Technology,Czech Republic Lagrange Multplers as Quanttatve Indcators n Economcs Ivan Mezník Insttute of Informatcs, Faculty of Busness and Management, Brno Unversty of TechnologCzech Republc Abstract The quanttatve role of Lagrange

More information

Luby s Alg. for Maximal Independent Sets using Pairwise Independence

Luby s Alg. for Maximal Independent Sets using Pairwise Independence Lecture Notes for Randomzed Algorthms Luby s Alg. for Maxmal Independent Sets usng Parwse Independence Last Updated by Erc Vgoda on February, 006 8. Maxmal Independent Sets For a graph G = (V, E), an ndependent

More information

Redesign of a University Hospital Preanesthesia Evaluation Clinic. using a Queuing Theory Approach

Redesign of a University Hospital Preanesthesia Evaluation Clinic. using a Queuing Theory Approach Reesgn of a Unvesty Hosptal Peanesthesa valuaton Clnc usng a Queung Theoy Appoach Maatje. Zonelan Mc ab Fe Boe MD PhD c Rcha J. Bouchee PhD b Annemek e Rooe MD PhD Jack W. van Kleef MD PhD a Dvson I Leen

More information

Concept and Experiences on using a Wiki-based System for Software-related Seminar Papers

Concept and Experiences on using a Wiki-based System for Software-related Seminar Papers Concept and Expeiences on using a Wiki-based System fo Softwae-elated Semina Papes Dominik Fanke and Stefan Kowalewski RWTH Aachen Univesity, 52074 Aachen, Gemany, {fanke, kowalewski}@embedded.wth-aachen.de,

More information

Reduced Pattern Training Based on Task Decomposition Using Pattern Distributor

Reduced Pattern Training Based on Task Decomposition Using Pattern Distributor > PNN05-P762 < Reduced Patten Taining Based on Task Decomposition Using Patten Distibuto Sheng-Uei Guan, Chunyu Bao, and TseNgee Neo Abstact Task Decomposition with Patten Distibuto (PD) is a new task

More information

Software Engineering and Development

Software Engineering and Development I T H E A 67 Softwae Engineeing and Development SOFTWARE DEVELOPMENT PROCESS DYNAMICS MODELING AS STATE MACHINE Leonid Lyubchyk, Vasyl Soloshchuk Abstact: Softwae development pocess modeling is gaining

More information

High Availability Replication Strategy for Deduplication Storage System

High Availability Replication Strategy for Deduplication Storage System Zhengda Zhou, Jingli Zhou College of Compute Science and Technology, Huazhong Univesity of Science and Technology, *, zhouzd@smail.hust.edu.cn jlzhou@mail.hust.edu.cn Abstact As the amount of digital data

More information

PAS: A Packet Accounting System to Limit the Effects of DoS & DDoS. Debish Fesehaye & Klara Naherstedt University of Illinois-Urbana Champaign

PAS: A Packet Accounting System to Limit the Effects of DoS & DDoS. Debish Fesehaye & Klara Naherstedt University of Illinois-Urbana Champaign PAS: A Packet Accountng System to Lmt the Effects of DoS & DDoS Debsh Fesehaye & Klara Naherstedt Unversty of Illnos-Urbana Champagn DoS and DDoS DDoS attacks are ncreasng threats to our dgtal world. Exstng

More information

AN IMPLEMENTATION OF BINARY AND FLOATING POINT CHROMOSOME REPRESENTATION IN GENETIC ALGORITHM

AN IMPLEMENTATION OF BINARY AND FLOATING POINT CHROMOSOME REPRESENTATION IN GENETIC ALGORITHM AN IMPLEMENTATION OF BINARY AND FLOATING POINT CHROMOSOME REPRESENTATION IN GENETIC ALGORITHM Main Golub Faculty of Electical Engineeing and Computing, Univesity of Zageb Depatment of Electonics, Micoelectonics,

More information

A Comparative Analysis of Data Center Network Architectures

A Comparative Analysis of Data Center Network Architectures A Compaative Analysis of Data Cente Netwok Achitectues Fan Yao, Jingxin Wu, Guu Venkataamani, Suesh Subamaniam Depatment of Electical and Compute Engineeing, The Geoge Washington Univesity, Washington,

More information

An Epidemic Model of Mobile Phone Virus

An Epidemic Model of Mobile Phone Virus An Epidemic Model of Mobile Phone Vius Hui Zheng, Dong Li, Zhuo Gao 3 Netwok Reseach Cente, Tsinghua Univesity, P. R. China zh@tsinghua.edu.cn School of Compute Science and Technology, Huazhong Univesity

More information

Scheduling Hadoop Jobs to Meet Deadlines

Scheduling Hadoop Jobs to Meet Deadlines Scheduling Hadoop Jobs to Meet Deadlines Kamal Kc, Kemafo Anyanwu Depatment of Compute Science Noth Caolina State Univesity {kkc,kogan}@ncsu.edu Abstact Use constaints such as deadlines ae impotant equiements

More information

Simultaneous Detection and Estimation, False Alarm Prediction for a Continuous Family of Signals in Gaussian Noise

Simultaneous Detection and Estimation, False Alarm Prediction for a Continuous Family of Signals in Gaussian Noise Sultaneous Detecton and Estaton, False Ala Pedcton fo a Contnuous Faly of Sgnals n Gaussan Nose D Mchael Mlde, Robet G Lndgen, and Mos M Bean Abstact New pobles ase when the standad theoy of jont detecton

More information

FXA 2008. Candidates should be able to : Describe how a mass creates a gravitational field in the space around it.

FXA 2008. Candidates should be able to : Describe how a mass creates a gravitational field in the space around it. Candidates should be able to : Descibe how a mass ceates a gavitational field in the space aound it. Define gavitational field stength as foce pe unit mass. Define and use the peiod of an object descibing

More information

A Lyapunov Optimization Approach to Repeated Stochastic Games

A Lyapunov Optimization Approach to Repeated Stochastic Games PROC. ALLERTON CONFERENCE ON COMMUNICATION, CONTROL, AND COMPUTING, OCT. 2013 1 A Lyapunov Optmzaton Approach to Repeated Stochastc Games Mchael J. Neely Unversty of Southern Calforna http://www-bcf.usc.edu/

More information

AN APPOINTMENT ORDER OUTPATIENT SCHEDULING SYSTEM THAT IMPROVES OUTPATIENT EXPERIENCE

AN APPOINTMENT ORDER OUTPATIENT SCHEDULING SYSTEM THAT IMPROVES OUTPATIENT EXPERIENCE AN APPOINTMENT ORDER OUTPATIENT SCHEDULING SYSTEM THAT IMPROVES OUTPATIENT EXPERIENCE Yu-L Huang Industral Engneerng Department New Mexco State Unversty Las Cruces, New Mexco 88003, U.S.A. Abstract Patent

More information

ON THE (Q, R) POLICY IN PRODUCTION-INVENTORY SYSTEMS

ON THE (Q, R) POLICY IN PRODUCTION-INVENTORY SYSTEMS ON THE R POLICY IN PRODUCTION-INVENTORY SYSTEMS Saifallah Benjaafa and Joon-Seok Kim Depatment of Mechanical Engineeing Univesity of Minnesota Minneapolis MN 55455 Abstact We conside a poduction-inventoy

More information

IBM Research Smarter Transportation Analytics

IBM Research Smarter Transportation Analytics IBM Reseach Smate Tanspotation Analytics Laua Wynte PhD, Senio Reseach Scientist, IBM Watson Reseach Cente lwynte@us.ibm.com INSTRUMENTED We now have the ability to measue, sense and see the exact condition

More information

The transport performance evaluation system building of logistics enterprises

The transport performance evaluation system building of logistics enterprises Jounal of Industial Engineeing and Management JIEM, 213 6(4): 194-114 Online ISSN: 213-953 Pint ISSN: 213-8423 http://dx.doi.og/1.3926/jiem.784 The tanspot pefomance evaluation system building of logistics

More information

An MILP model for planning of batch plants operating in a campaign-mode

An MILP model for planning of batch plants operating in a campaign-mode An MILP model for plannng of batch plants operatng n a campagn-mode Yanna Fumero Insttuto de Desarrollo y Dseño CONICET UTN yfumero@santafe-concet.gov.ar Gabrela Corsano Insttuto de Desarrollo y Dseño

More information