Achieving Energy Proportionality In Server Clusters

Size: px
Start display at page:

Download "Achieving Energy Proportionality In Server Clusters"

Transcription

1 Achevng Energy Proportonalty In Server Clusters Xnyng Zheng Ph.D Canddate /Electrcal and Computer Engneerng Mchgan Technologcal Unversty Houghton, 49931, US Yu Ca Assocate Professor /School of Technology Mchgan Technologcal Unversty Houghton, 49931, US Abstract Green computng s a hot ssue that has receved a great amount of nterests n the past few years. Energy proportonalty s a prncpal to ensure that energy consumpton s proportonal to the system workload. Energy proportonal desgn can effectvely mprove energy effcency of computng systems. In ths paper, an energy proportonal model s proposed based on queung theory and servce dfferentaton n server clusters, whch can provde controllable and predctable quanttatve control over power consumpton wth theoretcally guaranteed servce performance. Further study for the transton overhead s carred out correspondng strategy s proposed to compensate the performance degradaton caused by transton overhead. The model s evaluated va extensve smulatons and justfed by the real workload data trace. The results show that our model can acheve satsfed servce performance whle stll preservng energy effcency n the system. Keywords: green computng, energy proportonal, performance dfferentaton, transton overhead 1. INTRODUCTION Green computng s to support personal and busness computng needs n a green and sustanable manner, such as mnmzng stran and mpact on resources and envronment. Computng systems, partcularly enterprse data centers and hgh-performance cluster systems consume a sgnfcant amount of energy, thus placng an ncreasng burden on power supply and operatonal cost. For example, the power consumpton of enterprse data centers n the U.S. doubled between 2000 and 2005, and wll lkely trple agan n a few years [1]. In 2005, US data centers consumed 45 bllon kwh, whch was roughly 1.2 percent of the total amount of US electrcty consumpton, resultng n utlty blls of $2.7 bllon [2]. In 2006, the U.S. Congress passed blls to rase the IT ndustry s role n energy and envronmental polcy to the natonal level [3]. Furthermore, t s estmated that servers consume 0.5 percent of the world s total electrcty [4], whch f current demand contnues, s projected to quadruple by Some analysts predcted that IT nfrastructure power usage wll soon cost more than the hardware tself [5]. Many of the exstng works on power management n server clusters rely heavly on heurstcs or feedback control [6][7][8][9]. An mportant prncple n green computng s to ensure energy Internatonal Journal of Computer Networks (IJCN), Volume (1): Issue (2) 21

2 consumpton proportonalty, whch states that the energy consumpton should be proportonal to the system workload [10]. For example, when there s no or lttle workload, the system should consume no or lttle energy; when workload ncreases, energy consumpton should ncrease proportonally, untl the system reaches the full workload. Ths dea can effectvely mprove the energy effcency n real-lfe usage. Energy proportonalty s relatvely hard to be acheved on a standalone server because of hardware constrants. However, t s possble to acheve energy proportonalty on a server cluster, snce we can control the number of actve and nactve nodes n a server cluster. In ths paper, we propose an energy proportonal model n a server cluster and study ts performance n both sngle and multple classes scenaros. We further nvestgate the transton overhead based on ths model. The smulaton results show that the energy proportonal model can provde controllable and predctable quanttatve control over power consumpton wth theoretcally guaranteed servce performance. The rest of the paper s organzed as follows. Secton 2 revews related work. Secton 3 ntroduces the energy proportonal model. Performance metrcs and servers allocaton strategy are ntroduced n secton 4. Secton 5 evaluates the model and dscusses the transton overhead nfluence, a strategy s also proposed to compensate the transton overhead n ths secton, the model s further evaluated based on the real workload data trace, and the last secton concludes the paper. 2. RELATED WORK In lteratures, green computng s often related to terms lke green IT, sustanable computng, energy effcency, energy savng, power aware, power savng, and energy proportonal. In ths secton, we revew relevant technques commonly used on sngle server and server clusters. A. Sngle Server The green computng technques for a sngle server focus on mcroprocessors, memores and dsks. Current mcroprocessors allow power management by dynamc voltage and frequency scalng (DV/FS). DV/FS works because reducng the voltage and frequency provdes substantal savngs n power at the cost of slower program executon. Some researches te the scheduler drectly to DV/FS [11][12][13]. Most works deal exclusvely wth meetng real-tme schedulng deadlnes whle conservng energy. Tradtonally, many power management solutons rely heavly on heurstcs. Recently, feedback control theoretcal approaches for energy effcency have been proposed by a number of researchers. On a sngle server, recent works [14][15] proposed power control schemes based on feedback control theory. Femal et al. [16] developed an algorthm based on lnear programmng. In [8], a control theoretcal power management scheme on standalone servers was proposed. The feedback control theory s better than the tradtonal technques by provdng hgh accuracy and stablty. Thermal management s another ssue n power-aware computng, snce temperature s a byproduct of power dsspaton [17]. Recent research demonstrated that dynamc thermal management (DTM) can respond to thermal condtons by adaptvely adjustng a chp power consumpton profle on the accordng to feedback from temperature sensors [14] [18]. Research work on memory s often combned wth processors and dsks. In [19], the authors used open-loop control to shft power between processor and memory to mantan a server power budget. In [20], they proposed a soluton to store pages and relablty data n dle RAM nstead of usng slow dsk. A large porton of the power budget of servers goes nto the I/O subsystem, the dsk array n partcular. Many dsk systems offer multple power modes and can be swtched to a low power mode when not n use to acheve energy savng. Such technques had been proposed Internatonal Journal of Computer Networks (IJCN), Volume (1): Issue (2) 22

3 n [21][22]. Sudhanva et al. [23] presented a new approach called DRPM to modulate dsk speed dynamcally, and a practcal mplementaton was provded for ths mechansm. B. Server Clusters In recent years, power management has become one of the most mportant concerns on server clusters. Some methods proposed on a sngle server can be extended to server clusters. In [24][25], the authors presented smlar ways of applyng DV/FS and cluster reconfguraton, usng threshold values, based on the utlzaton of the system load to keep the processor frequences as low as possble, wth less actve nodes. In [9], the authors extended the feedback control scheme to clusters. Power has been used as a tool for applcaton-level performance requrements. Sharma et al. [26] proposed feedback control schemes to control applcaton-level qualty of servce requrements. Chen et al. [27] presented a feedback controller to manage the response tme n server clusters. Some researchers appled DTM on an entre data center rather than ndvdual servers or chps. In [28], the authors lad out polces for workload placement to promote unform temperature dstrbuton usng actve thermal zones. Vary-On Vary-off (VOVF) s a dynamc structure confguraton mechansm to ensure energyaware computng n server clusters, whch turns nodes on and off to adjust the number of actve servers by the workload. Other work had been carred out based on VOVF [29][30][28]. In [31], The authors proposed a method to reduce network energy consumpton va sleepng and rate adaptaton by combnng VOVF and DV/FS. Another group developed power savng technques for connecton orented servers [32]. The authors tested server provsonng and load dspatchng on the MSN nstant messagng framework, and evaluated varous load skewng technques to trade off energy savng and qualty of servce. Vrtualzaton s another key strategy to reduce power consumpton n enterprse networks. Wth vrtualzaton, multple vrtual servers can be hosted on less but more powerful physcal servers, usng less electrcty [33]. In [34], researchers developed methods to effcently manage the aggregate platform resources accordng to the guest vrtual machnes (VM) of relatve mportance (Class-of-Servce), usng both the black-box and the VM-specfc approach. Hu et al. [35] used lve mgraton of vrtual machnes to transfer load among the nodes on a multlayer rng-based overlay. In [4], researchers scheduled vrtual machnes n a computer cluster to reduce power consumpton va the technque of Dynamc Voltage Frequency Scalng (DVFS). An economy drven energy and resource management framework was presented for clusters n [36]. Each servce bds for resources as a functon of delvered performance. In [37], researchers formulated the problem as a cooperatve game, and used game theory to fnd the barganng pont. The energy-related budget has accounted for a large porton of total storage system cost of ownershp. Some studes tred multspeed dsks for servers [23][38]. Other technques were ntroduced to regulate data movement. For example, the mostly used data can be transferred to specfc dsks or memory, thus other dsks can be set to a low power mode [39]. 3. ENERGY PROPORTIONAL MODEL A. Energy Proportonal Model on a Sngle Server The energy proportonal model states that energy consumpton P should be proportonal to the workload λ, whle ensurng servce performance. P = a λ + b (1) Internatonal Journal of Computer Networks (IJCN), Volume (1): Issue (2) 23

4 Fg. 1. The energy consumpton curves of non-energy proportonal server and strct energy proportonal server. Fgure.1 conceptually llustrates the energy consumpton curve n non-energy proportonal servers and energy proportonal servers. The typcal server operatng range s between 10% - 60%. We can see that n a non-energy proportonal server, t stll consumes about half of ts full power when dong vrtually no work [10]. Energy proportonal server deally consumes no power when dle (b = 0), nearly no power when very lttle work s performed, and gradually more power as the actvty level ncreases. Energy-proportonal desgns would enable large energy savngs on servers. However, most servers nowadays are CPU, memory and hard dsk ntensve servers. The energy consumpton of CPU s almost lnear to ts utlzaton [32]. But memory and hard dsks are nonlnear energy consumpton components. As a result, energy proportonalty s not easy to be acheved on a standalone sever because of the hardware constrants. B. Energy Proportonal Model on Server Clusters It s more feasble to acheve energy proportonalty n a server cluster. Most computng systems nowadays have at least two modes of operaton: an actve mode when the system s workng and an dle mode when the system s nactve and consumes lttle energy. Some researchers proposed to have fner-graned power modes, runnng at low speed and wth lower power supply voltage. It s known that to quanttatvely control energy consumpton, one feasble way s to adaptvely and dynamcally control the number of servers runnng n actve and nactve modes accordng to system workload. For smplcty, we assume all the servers n the cluster are dentcal nodes. On typcal web servers and web clusters, system workload can be descrbed by the request arrval rate λ. Let M be the total number of servers n the cluster, and Λ be the maxmum arrval rate for the cluster. m s the total number of actve servers. The total energy consumpton of a server cluster s: ac ( ) n (2) P = m P + M m P Pac s the power consumpton of fully actve nodes; n P s the power consumpton of nactve nodes. Based on the energy proportonal model, we have: P P max λ = Λ r (3) where P max = M P ac. r s a parameter, whch adjusts the energy consumpton curve n Fgure 1. The ratonale of usng parameter r s as follows. Ideally the r s set to r =1 where energy consumpton s strctly proportonal to workload. However, we can adjust t to satsfy dfferent Internatonal Journal of Computer Networks (IJCN), Volume (1): Issue (2) 24

5 performance constrants. Wth the help of (2), we can rewrte equaton (3) as: Pac m = ( r M Pn ) / ( Pac Pn ) (4) Λ M Λ / M Here s the maxmum jobs that a sngle cluster node can handle. Ideally P n = 0, whch ndcates that a server consumes no energy when t s runnng on an nactve mode. For smplcty, we suppose P n = 0 n ths paper, ths assumpton wll not affect the performance of our model. We fnally acheve that the total number of actve servers m s determned by the system workload λ : λ m = r Λ M (5) The number of servers may not be an nteger based on (5). We wll set the nteger no less than m, whch s the mnmal number of servers to run n fully actve mode. 4. SERVERS ALLOCATION BASED ON ENERGYPROPORTIONAL MODEL An mportant task of energy aware computng s to acheve energy effcency whle ensurng performance. In ths secton, we wll descrbe how to allocate servers accordng to workload, whle ensurng qualty of servces (QoS) metrcs. A. Performance Metrcs One mportant and commonly used QoS metrc on Internet servers s slowdown, whch s defned as the dvson of watng tme by servce tme. Another commonly used performance metrc s request tme whch s the sum of watng tme and servce tme. We choose slowdown and request tme as performance metrcs n our model because they are related to both watng tme and servce tme. Our theoretcal framework s bult along the lne of the prevous servce dfferentaton models presented n [40][41][42][43]. In our network model, a heavy-taled dstrbuton of packet sze s used to descrbe web traffc. Here we assume that the servce tme s proportonal to the packet sze. The packet nter-arrval tme follows exponental dstrbuted wth a mean of 1/λ, where λ s the arrval rate of ncomng packets. A set of tasks wth sze followng a heavy-taled Bounded Pareto dstrbuton are characterzed by three parameters: α the shape parameter; k, the shortest possble job; p, the upper bound of jobs. The probablty densty functon can be defned as: 1 α α 1 f ( x) k x α 1 ( k p) α = (6) where, α, k > 0, k x p. If we defne a functon: then we have: K αk α ( α, k, p) = α 1 ( k p) (7) Internatonal Journal of Computer Networks (IJCN), Volume (1): Issue (2) 25

6 K ( α, k, p) K ( α 1, k, p) ( ) ( α ) p f α 1; E[ X ] = f ( x) dx = k ln p ln k K, k, p f α = 1. (8) Smlarly, we can derve 2 E[ X ] and E[ X ] ( α k p) ( α 2, k, p) p 2 2 K,, E[ X ] = f ( x) x dx = (9) k K ( α k p) ( α + 1, k, p) p 1 1 K,, E[ X ] = f ( x) x dx = (10) k K Accordng to Pollaczek-Khnchn formula, the average watng tme for the ncomng packets s: 2 λe[ X ] E[ W ] = 2 1 [ ] ( λe X ) We can derve a closed-form expresson of the expected slowdown n an M/G/1 queue on a sngle Internet server. E[ S] = E[ W ] E[ X ] = The expected request tme wth the ncomng job rate s: λe[ X ] E[ X ] ( λe X ) 2 1 [ ] 2 λe[ X ] E[ R] = E[ W ] + E[ X ] = + E[ X ] 2 1 [ ] ( λe X ) B. Servers Allocaton for a Sngle Class In ths secton, we assume all the ncomng requests are classfed nto just one class. We want to ensure the QoS metrcs based on dfferent workload λ. For example, the expected request tme of the ncomng jobs should stay wthn a threshold, E[ R] < β. We assume m (11) (12) (13) s the number of actve server nodes handlng the ncomng requests. When usng a round-robn dspatchng polcy, the packet arrval rate of a node s can be calculated as: λ m 2 λe[ X ] E[ R] = + E[ X ] < β 2 [ ] ( m λe X ). The expected request tme n a server cluster Based on the above energy proportonalty (5), equaton (14) can be re-wrtten as: 2 E[ X ] E[ R] = E[ X ] 2 [ ] + < β ( r M Λ E X ) It s easy to observe that request tme s not dependng on workload, we can just adjust parameter r to satsfy dfferent performance thresholds. C. Servers Allocaton on Servce Dfferentaton Now we study server allocaton schemes for servce dfferentaton. In a cluster system, the ncomng requests are often classfed nto N classes. Each class may have dfferent QoS requrements. We assume m s the number of actve server nodes n class, and λ s the arrval rate n class. The expected slowdown of class n a server cluster can be calculated as: (14) (15) Internatonal Journal of Computer Networks (IJCN), Volume (1): Issue (2) 26

7 λ E X E[ S ] = 2 [ ] 2 1 [ ] E[ X ] ( m λ E X ) Here we choose not to use request tme as a performance metrc for servce dfferentaton because of ts overly complcated mathematcal expresson. However, each class should satsfy the request tme constrant. Obvously the results presented n ths paper wll not be affected by the selecton of performance metrcs. We adopt a relatve servce dfferentaton model where the QoS factor of slowdown between dfferent classes are based on ther predefned dfferentaton parameters. E[ S ] E[ S ] Where 1, j N : We assume class 1 s the hghest class and set < 1 < 2 < L < receve better servce,.e., lower slowdown [39]. j δ δ (16) = (17) j δ δ δ 0 N, then hgher classes Based on the above energy proportonalty and servce dfferentaton model, accordng to formula (5)(17), we can derve the server allocaton scheme n a cluster system as N % M λ λ * [ ] = 1 r E X Λ m = λ E[ X ] + (18) N % Here m s the number of actve servers n class, and % λ = λ δ s the normalzed arrval rate. The frst term of formula (18) ensures that the sub-cluster n class wll not be overloaded. The second term s related to arrval rates, dfferentaton parameters, and r. We can also derve the expected slowdown of class as: 2 1 N δe[ X ] E[ X ] % λ = 1 E[ S ] = (19) N M 2 λ * [ ] = 1 r E X Λ From (19) we can observe that the slowdown of class s proportonal to the pre-specfed parameter δ, and s related to r. The slowdown rato only depends on the pre-defned dfferentaton parameters. The expected request tme for class can be calculated as: 2 N δe[ X ] % λ = 1 E[ R ] = + E[ X ] N M 2 λ * [ ] = 1 r E X Λ β s request tme constrant for class. We can learn from equaton (20), request tme n class s also ndependent of workload, but depends on both the pre-specfed parameter δ and r. The performance s controllable based on our energy proportonal model, wth acceptable performance degradaton; large amounts of energy can be saved. = 1 λ β (20) Internatonal Journal of Computer Networks (IJCN), Volume (1): Issue (2) 27

8 5. PERFORMANCE EVALUATION A. Smulaton Results We buld a smulator whch conssts of a package generator, a server dspatcher, a number of watng queues, and a number of servers. The package generator produces ncomng requests wth exponental nter-arrval tme dstrbuton and bounded Pareto packet sze dstrbuton. The GNU scentfc lbrary s used for stochastc smulaton. Smulaton parameters are set as follows. The shape parameter α of the bounded Pareto dstrbuton s set to 1.5. The lower bound k and upper bound p were set to 0.1 and 100, respectvely [44]. The number of servers n the cluster s 20. And we set the normalzed maxmum jobs one server can handle [32]. Λ M = 1. We set the power consumpton 160W for actve nodes Fg. 2. Comparson of request tme between non-energy proportonal model and energy Proportonal model. r s set dfferently accordng to dfferent requrements of performance n a sngle class scenaro. Fg. 3. Comparson of power consumpton between non-energy proportonal model and energy proportonal model n a sngle class scenaro. r s set dfferently accordng to dfferent requrements of performance. we can acheve consderable energy savng wth energy proportonal model. Internatonal Journal of Computer Networks (IJCN), Volume (1): Issue (2) 28

9 Fg. 4. Comparson of request tme n hgher prorty class between non-energy proportonal model and energy proportonal models. r s set dfferently accordng to dfferent requrements of performance n a multple classes scenaro. Fg. 5. Comparson of request tme n lower prorty class between Non-energy proportonal model and energy proportonal models. r s set dfferently accordng to dfferent requrements of performance n a multple classes scenaro. We frst evaluate the energy proportonal model for the sngle class scenaro. We set the request tme β= 0.9, β= 1.1 and β= 1.3 whch correspond to adjustment parameter r = 1.1, r = 1 and r = 0.9 respectvely. We show the smulaton results n the workload range of 10% - 80%. When the workload s above 80%, the mpact of energy proportonalty constrant s very lmted. Snce the typcal server operatng range s between 10% - 60%, the results presented here are suffcent to test the energy proportonal model. As Fgure 2 ndcates that the request tme s always around the pre-defned performance parameter under dfferent workload. The request tme ncreases as the value of r decreases. The results show that wth adjustable parameter r desrable servce performance can be acheved. Fgure 3 compares the energy consumpton of energy proportonal model and non-energy proporton model for a sngle class scenaro. We can acheve better energy effcency under low workload, whch leads to large amounts of energy savng n a server cluster. Internatonal Journal of Computer Networks (IJCN), Volume (1): Issue (2) 29

10 Next, we compare the performance metrcs n a multple classes scenaro, as shown n Fgure 4, 5, 6. The number of classes s normally two or three [45][46]. In ths paper we choose two classes of ncomng requests. We set the target slowdown rato δ2 : δ 1 = 2 : 1. The energy curve parameters are set dfferently accordng to dfferent request tme constrants. Note, n a multple classes scenaro, parameter r s determned by performance requrements of all classes, whch means t should be set to be the largest value satsfyng the requrements of all the classes. We observe that the model can acheve desrable proportonalty of slowdown dfferentaton wth request tme constrants. Fgure 7 also compares the energy consumptons for proportonal and non-proportonal models n multple classes scenaro. Fg. 6. Comparson of slowdown rato between non-energy proportonal model and energy proportonal models. r s set by dfferent requrements of performance n a multple classes scenaro. Fg. 7. Comparson of power consumpton between non-energy proportonal model and energy proportonal model n multple classes scenaro. r s set by dfferent requrements of performance. We can acheve consderable energy savng compare to the non-energy proportonal model. B. Transton Overhead Analyss The model proposed n ths paper s a contnual allocaton process, where we dynamcally change the number of actve servers. The transton tme when a server transfers from an nactve mode to an actve mode can not be gnored, ths can nfluence the performance durng the transton perod. Thus, t s necessary to estmate the cost of transton overhead. Internatonal Journal of Computer Networks (IJCN), Volume (1): Issue (2) 30

11 Generally speakng, the transton tme for dfferent servers s dfferent whch depends on the processor and other hardware constrants. Therefore, we study the nfluence on performance caused by transton overhead under dfferent tme. Fgure 8 shows how the request tme changes when consderng transton overhead as the workload gradually changed from 0%-80% based on the energy proportonal model. We only concern the stuaton when the workload ncreases, snce as the workload decreases, the number of actve servers wll declne, whch wll not cause performance degradaton. The y-axs s the request tme under dfferent transton overhead. As ndcated n the fgure, larger transton tme has more mpact on performance. The performance wll be affected greatly when large number of servers can not transfer to actve mode on tme. Fg. 8. The effect to performance of transton overhead n energy proportonal model, the transton tme s set to be 15,20,25,30 respectvely Fg. 9. Request tme after addng one spare server based on energy proportonal model n a sngle class scenaro, the transton tme s set to be 15,20,25,30 respectvely. It s mportant to make sure that the QoS s not sacrfced excessvely n favor of power and energy savngs. Spare servers are added to solve the problem of transton overhead. Fgure 9, 10 llustrate the performance after one and two spare servers are added n a sngle class scenaro. By addng one spare server, the performance can be mproved dramatcally compared to the case of no spare server. Addng two spare servers, the response tme can stay under the pre-defned threshold when the workload gradually changes from 0%-80%. However, n some specal stuatons, the workload may vary sgnfcantly wthn two control perods. One or two spare servers are not adequate to compensate the performance degradaton. More spare servers Internatonal Journal of Computer Networks (IJCN), Volume (1): Issue (2) 31

12 are requred. C. Performance Evaluaton Based on Real Workload Data Trace To evaluate the model on realstc traffc patterns, we use an hour s workload trace collected by Lawrence Berkeley Natonal Laboratory [47]. Request tme threshold s set to be β= 0.6 and r = 1. Fgure 11 llustrates the performance based on our model n a sngle class scenaro. The requests arrval rate and job sze are normalzed. We evaluate the performance n the stuatons of non-spare server and spare servers respectvely. As shown n the fgure, when the workload decreases, there s no performance degradaton, however the performance degradaton can be clearly seen as the workload ncreases n the case of no spare server s added. Wth one or two spare servers, the performance can be mproved sgnfcantly. Especally, when two spare servers are always on, request tme s always under predefned threshold. The result also ndcates that as the number of spare server ncreases, the performance does not change dramatcally. The request tme tends to stay n a level, whch demonstrates proper spare servers should be set to compensate the performance degradaton. Fg. 10. Request tme when addng two spare servers based on energy proportonal model n a sngle class scenaro, the transton tme s set to be 15,20,25,30 respectvely. Fg. 11. Request tme when addng two spare servers based on energy proportonal model n a sngle class scenaro. Internatonal Journal of Computer Networks (IJCN), Volume (1): Issue (2) 32

13 Fg. 12. Power consumpton when addng two spare servers based on energy proportonal model n a sngle class scenaro. Fgure 12 evaluates the power consumpton based on our model under real workload data trace. The system arrval rate s the same as shown n fgure 11. The power consumpton s dynamcally changed as the workload changed. Wth lttle more power consumpton, we can acheve better performance, and elmnate the effect of transton overhead. 6. CONCLUSION AND FUTURE WORK Energy management becomes a key ssue n server clusters and data centers. Ths paper ams at provdng effectve strateges to reduce power consumpton and reduce the mpact of performance degradaton. We summarze out work as follows: frst, the energy proportonal model based on queung theory can provde accurate, controllable and predctable quanttatve control over power consumpton; second, we analyze the effect of transton overhead and propose a strategy to mprove the performance effcency. Fnally we evaluate the energy proportonal model va smulaton. Smulaton results show that the energy proportonal model can acheve predctable and controllable proportonal energy consumpton and desrable performance n a server cluster. Future work would nclude studyng the effect on performance when applyng dfferent dspatchng strateges n our model. We are stll tryng to extend the server states to solve the problem of nonnteger number of nodes, whch wll further enhance the energy effcency. Eventually our goal s to apply our model to the real Internet web servers n the future. 7. REFERENCES [1] U.S. Envronmental Protecton Agency. Report to Congress on Server and Data Center Energy Effcency.August [2] J. S. Aronson, Makng t a postve force n envronmental change, IT Professonal, vol. 10, pp , Jan [3] US Congress. House bll To study and promote the use of energy efcent computer servers n the unted states. Retreved: [4] G. von Laszewsk, L. Wang, A. J. Younge, and X. He, Poweraware schedulng of vrtual machnes n dvfs enabled clusters, Cluster Computng and Workshops, CLUSTER 09. IEEE Internatonal Conference on, pp. 1 10, Jan [5] Y. Chen, A. Das, W. Qn, A. Svasubramanam, and Q. Wang, Managng server energy and operatonal costs n hostng centers, Proceedngs of the 2005 ACM SIGMETRICS nternatonal, Jan Internatonal Journal of Computer Networks (IJCN), Volume (1): Issue (2) 33

14 [6] C. Lefurgy, K. Rajaman, F. Rawson, and W. Felter, Energy management for commercal servers, Computer, Jan [7] Y. Lu and G. D. Mchel, Operatng-system drected power reducton, In proc. of nternatonal symposum on Low power electroncs and desgn, Jan [8] C. Lefurgy, X. Wang, and M. Ware, Server-level power control, Autonomc Computng, ICAC 07. Fourth Internatonal Conference on, pp. 4 4, May [9] X. Wang and M. Chen, Cluster-level feedback power control for performance optmzaton, In Proc. of Symposum on Hgh-Performance Computer Archtecture, Jan [10] L. Barroso and U. Holzle, The case for energy-proportonal computng, Computer, vol. 40, pp , Dec [11] G. Quan and X. Hu, Energy effcent fxed-prorty schedulng for real-tme systems on varable voltage processors, Desgn Automaton Conference, Jan [12] M. Elnozahy, M. Kstler, and R. Rajamony, Energy conservaton polces for web servers, Proceedngs of the 4th conference on USENIX Symposum on Internet Technologes and Systems, Jan [13] J. Pouwelse, K. Langendoen, and H. Sps, Energy prorty schedulng for varable voltage processors, Proceedngs of the 2001 nternatonal symposum on Low power, Jan [14] K. Skadron, T. Abdelzaher, and M. Stan, Control-theoretc technques and thermal-rc modelng for accurate and localzed dynamc thermal management, pp , Feb [15] Q. Wu, P. Juang, M. Martonos, L. Peh, and D. Clark, Formal control technques for powerperformance management, IEEE Mcro, Jan [16] M. Femal and V. Freeh, Boostng data center performance through non-unform power allocaton, Autonomc Computng, Jan [17] R. Graybll and R. Melhem, Power aware computng, books.google.com, Jan [18] D. Brooks and M. Martonos, Dynamc thermal management for hgh-performance mcroprocessors, Hgh-Performance Computer Archtecture, Jan [19] W. Felter, K. Rajaman, T. Keller, and C. Rusu, A performanceconservng approach for reducng peak power consumpton n server systems, Proceedngs of the 19th annual nternatonal conference on Supercomputng, Jan [20] T. Newhall, D. Amato, and A. Pshenchkn, Relable adaptable network ram, 2008 IEEE Internatonal Conference on Cluster Computng, Jan [21] A. Wessel, B. Beutel, and F. Bellosa, Cooperatve I/O a novel I/O semantcs for energyaware applcatons, usenx.org. [22] D. Helmbold, D. Long, T. Sconyers, and B. Sherrod, Adaptve dsk spndown for moble computers, Moble Networks and Applcatons, Jan [23] S. Gurumurth, A. Svasubramanam, and M. Kandemr, DRPM: dynamc speed control for power management n server class dsks, Computer Archtecture, Jan [24] M. Vasc, O. Garca, J. Olver, P. Alou, and J. Cobos, A dvs system based on the trade-off between energy savngs and executon tme, Control and Modelng for Power Electroncs, COMPEL th Workshop on, pp. 1 6, Jul [25] E. Pnhero, R. Banchn, E. Carrera, and T. Heath, Dynamc cluster reconfguraton for power and performance, Complers and operatng systems for low power, Jan [26] V. Sharma, A. Thomas, T. Abdelzaher, K. Skadron, and Z. Lu, Poweraware qos management n web servers, 24th IEEE Real-Tme Systems Symposum, Jan [27] C. Dovrols, D. Stlads, and P. Ramanathan, Proportonal dfferentated servces: Delay dfferentaton and packet schedulng, Proceedngs of the conference on Applcatons, Jan [28] R. Sharma, C. Bash, C. Patel, and R. Fredrch, Balance of power: Dynamc thermal management for nternet data centers, IEEE Internet Computng, Jan [29] X. Fan, W. Weber, and L. Barroso, Power provsonng for a warehouse-szed computer, Proceedngs of the 34th annual nternatonal conference on archtecture, Jan B [30] R. Guerra, J. Lete, and G. Fohler, Attanng soft real-tme constrant and energy-effcency n web servers, Proceedngs of the 2008 ACM symposum on Appled computng, Jan [31] S. Nedevsch, L. Popa, G. Iannaccone, and S. Ratnasamy, Reducng network energy consumpton va sleepng and rate-adaptaton, NSDI, Jan [32] G. Chen, W. He, J. Lu, S. Nath, L. Rgas, and L. Xao, Energy-aware server provsonng and load dspatchng for connecton-ntensve nternet servces, Proceedngs of the 5th USENIX Internatonal Journal of Computer Networks (IJCN), Volume (1): Issue (2) 34

15 Symposum on Networked Systems Desgn and Implementaton, Jan [33] S. Murugesan, Harnessng green t: Prncples and practces, IT Professonal, Jan [34] M. Kesavan, A. Ranadve, A. Gavrlovska, and K. Schwan, Actve coordnaton (act) toward effectvely managng vrtualzed multcore clouds, 2008 IEEE Internatonal Conference on Cluster Computng, Jan [35] L. Hu, H. Jn, X. Lao, X. Xong, and H. Lu, Magnet: A novel schedulng polcy for power reducton n cluster wth vrtual machnes, 2008 IEEE Internatonal Conference on Cluster Computng, Jan [36] J. Chase, D. Anderson, P. Thakar, and A. Vahdat, Managng energy and server resources n hostng centers, Proceedngs of the eghteenth ACM symposum on Operatng Operatng System Prncples, Jan [37] I. Ahmad, S. Ranka, and S. Khan, Usng game theory for schedulng tasks on mult-core processors for smultaneous optmzaton of performance and energy, pp. 1 6, Aprl [38] E. Carrera, E. Pnhero, and R. Banchn, Conservng dsk energy n network servers, Proceedngs of the 17th annual nternatonal conference on Supercomputng, Jan [39] M. Song, Energy-aware data prefetchng for mult-speed dsks n vdeo servers, Proceedngs of the 15th nternatonal conference on Supercomputng, Jan [40] X. Zhou, Y. Ca, C. Chow, and M. Augustejn, Two-ter resource allocaton for slowdown dfferentaton on server clusters, Parallel Processng, Jan [41] X. Zhou, Y. Ca, G. Godavar, and C. Chow, An adaptve process allocaton strategy for proportonal responsveness dfferentaton on web servers, Web Servces, Proceedngs. IEEE Internatonal Conference on, pp , Jun [42] C. Dovrols and P. Ramanathan, A case for relatve dfferentated servces and the proportonaldfferentaton model, Network, Jan [43] X. Zhou and C. Xu, Harmonc proportonal bandwdth allocaton and schedulng for servce dfferentaton on streamng servers, Parallel and Dstrbuted Systems, IEEE Transactons on, vol. 15, pp , Sep [44] M. Harchol-Balter and C. U. PITTSBURGH, Task assgnment wth unknown duraton, do.eeecomputersocety.org, Jan [45] H. Zhu, H. Tang, and T. Yang;, Demand-drven servce dfferentaton n cluster-based network servers, INFOCOM Twenteth Annual Jont Conference of the IEEE Computer and Communcatons Socetes. Proceedngs. IEEE, vol. 2, pp vol.2, Mar [46] L. Zhang, A two-bt dfferentated servces archtecture for the nternet, Request for Comments (Informatonal), Jan [47] NASA Kennedy Space Center Server Traces. Internatonal Journal of Computer Networks (IJCN), Volume (1): Issue (2) 35

Parallelism for Nested Loops with Non-uniform and Flow Dependences

Parallelism for Nested Loops with Non-uniform and Flow Dependences Parallelsm for Nested Loops wth Non-unform and Flow Dependences Sam-Jn Jeong Dept. of Informaton & Communcaton Engneerng, Cheonan Unversty, 5, Anseo-dong, Cheonan, Chungnam, 330-80, Korea. seong@cheonan.ac.kr

More information

Simulation Based Analysis of FAST TCP using OMNET++

Simulation Based Analysis of FAST TCP using OMNET++ Smulaton Based Analyss of FAST TCP usng OMNET++ Umar ul Hassan 04030038@lums.edu.pk Md Term Report CS678 Topcs n Internet Research Sprng, 2006 Introducton Internet traffc s doublng roughly every 3 months

More information

Video Proxy System for a Large-scale VOD System (DINA)

Video Proxy System for a Large-scale VOD System (DINA) Vdeo Proxy System for a Large-scale VOD System (DINA) KWUN-CHUNG CHAN #, KWOK-WAI CHEUNG *# #Department of Informaton Engneerng *Centre of Innovaton and Technology The Chnese Unversty of Hong Kong SHATIN,

More information

Virtual Memory. Background. No. 10. Virtual Memory: concept. Logical Memory Space (review) Demand Paging(1) Virtual Memory

Virtual Memory. Background. No. 10. Virtual Memory: concept. Logical Memory Space (review) Demand Paging(1) Virtual Memory Background EECS. Operatng System Fundamentals No. Vrtual Memory Prof. Hu Jang Department of Electrcal Engneerng and Computer Scence, York Unversty Memory-management methods normally requres the entre process

More information

Virtual Machine Migration based on Trust Measurement of Computer Node

Virtual Machine Migration based on Trust Measurement of Computer Node Appled Mechancs and Materals Onlne: 2014-04-04 ISSN: 1662-7482, Vols. 536-537, pp 678-682 do:10.4028/www.scentfc.net/amm.536-537.678 2014 Trans Tech Publcatons, Swtzerland Vrtual Machne Mgraton based on

More information

Energy-Efficient Workload Placement in Enterprise Datacenters

Energy-Efficient Workload Placement in Enterprise Datacenters COVER FEATURE CLOUD COMPUTING Energy-Effcent Workload Placement n Enterprse Datacenters Quan Zhang and Wesong Sh, Wayne State Unversty Power loss from an unnterruptble power supply can account for 15 percent

More information

A mathematical programming approach to the analysis, design and scheduling of offshore oilfields

A mathematical programming approach to the analysis, design and scheduling of offshore oilfields 17 th European Symposum on Computer Aded Process Engneerng ESCAPE17 V. Plesu and P.S. Agach (Edtors) 2007 Elsever B.V. All rghts reserved. 1 A mathematcal programmng approach to the analyss, desgn and

More information

AADL : about scheduling analysis

AADL : about scheduling analysis AADL : about schedulng analyss Schedulng analyss, what s t? Embedded real-tme crtcal systems have temporal constrants to meet (e.g. deadlne). Many systems are bult wth operatng systems provdng multtaskng

More information

NUMERICAL SOLVING OPTIMAL CONTROL PROBLEMS BY THE METHOD OF VARIATIONS

NUMERICAL SOLVING OPTIMAL CONTROL PROBLEMS BY THE METHOD OF VARIATIONS ARPN Journal of Engneerng and Appled Scences 006-017 Asan Research Publshng Network (ARPN). All rghts reserved. NUMERICAL SOLVING OPTIMAL CONTROL PROBLEMS BY THE METHOD OF VARIATIONS Igor Grgoryev, Svetlana

More information

Real-Time Guarantees. Traffic Characteristics. Flow Control

Real-Time Guarantees. Traffic Characteristics. Flow Control Real-Tme Guarantees Requrements on RT communcaton protocols: delay (response s) small jtter small throughput hgh error detecton at recever (and sender) small error detecton latency no thrashng under peak

More information

Scheduling Remote Access to Scientific Instruments in Cyberinfrastructure for Education and Research

Scheduling Remote Access to Scientific Instruments in Cyberinfrastructure for Education and Research Schedulng Remote Access to Scentfc Instruments n Cybernfrastructure for Educaton and Research Je Yn 1, Junwe Cao 2,3,*, Yuexuan Wang 4, Lanchen Lu 1,3 and Cheng Wu 1,3 1 Natonal CIMS Engneerng and Research

More information

RAP. Speed/RAP/CODA. Real-time Systems. Modeling the sensor networks. Real-time Systems. Modeling the sensor networks. Real-time systems:

RAP. Speed/RAP/CODA. Real-time Systems. Modeling the sensor networks. Real-time Systems. Modeling the sensor networks. Real-time systems: Speed/RAP/CODA Presented by Octav Chpara Real-tme Systems Many wreless sensor network applcatons requre real-tme support Survellance and trackng Border patrol Fre fghtng Real-tme systems: Hard real-tme:

More information

Efficient Distributed File System (EDFS)

Efficient Distributed File System (EDFS) Effcent Dstrbuted Fle System (EDFS) (Sem-Centralzed) Debessay(Debsh) Fesehaye, Rahul Malk & Klara Naherstedt Unversty of Illnos-Urbana Champagn Contents Problem Statement, Related Work, EDFS Desgn Rate

More information

Evaluation of Parallel Processing Systems through Queuing Model

Evaluation of Parallel Processing Systems through Queuing Model ISSN 2278-309 Vkas Shnde, Internatonal Journal of Advanced Volume Trends 4, n Computer No.2, March Scence - and Aprl Engneerng, 205 4(2), March - Aprl 205, 36-43 Internatonal Journal of Advanced Trends

More information

A Model Based on Multi-agent for Dynamic Bandwidth Allocation in Networks Guang LU, Jian-Wen QI

A Model Based on Multi-agent for Dynamic Bandwidth Allocation in Networks Guang LU, Jian-Wen QI 216 Jont Internatonal Conference on Artfcal Intellgence and Computer Engneerng (AICE 216) and Internatonal Conference on etwork and Communcaton Securty (CS 216) ISB: 978-1-6595-362-5 A Model Based on Mult-agent

More information

An Optimal Algorithm for Prufer Codes *

An Optimal Algorithm for Prufer Codes * J. Software Engneerng & Applcatons, 2009, 2: 111-115 do:10.4236/jsea.2009.22016 Publshed Onlne July 2009 (www.scrp.org/journal/jsea) An Optmal Algorthm for Prufer Codes * Xaodong Wang 1, 2, Le Wang 3,

More information

Distributed Resource Scheduling in Grid Computing Using Fuzzy Approach

Distributed Resource Scheduling in Grid Computing Using Fuzzy Approach Dstrbuted Resource Schedulng n Grd Computng Usng Fuzzy Approach Shahram Amn, Mohammad Ahmad Computer Engneerng Department Islamc Azad Unversty branch Mahallat, Iran Islamc Azad Unversty branch khomen,

More information

Some material adapted from Mohamed Younis, UMBC CMSC 611 Spr 2003 course slides Some material adapted from Hennessy & Patterson / 2003 Elsevier

Some material adapted from Mohamed Younis, UMBC CMSC 611 Spr 2003 course slides Some material adapted from Hennessy & Patterson / 2003 Elsevier Some materal adapted from Mohamed Youns, UMBC CMSC 611 Spr 2003 course sldes Some materal adapted from Hennessy & Patterson / 2003 Elsever Scence Performance = 1 Executon tme Speedup = Performance (B)

More information

Load Balancing for Hex-Cell Interconnection Network

Load Balancing for Hex-Cell Interconnection Network Int. J. Communcatons, Network and System Scences,,, - Publshed Onlne Aprl n ScRes. http://www.scrp.org/journal/jcns http://dx.do.org/./jcns.. Load Balancng for Hex-Cell Interconnecton Network Saher Manaseer,

More information

Analysis of Collaborative Distributed Admission Control in x Networks

Analysis of Collaborative Distributed Admission Control in x Networks 1 Analyss of Collaboratve Dstrbuted Admsson Control n 82.11x Networks Thnh Nguyen, Member, IEEE, Ken Nguyen, Member, IEEE, Lnha He, Member, IEEE, Abstract Wth the recent surge of wreless home networks,

More information

Mathematics 256 a course in differential equations for engineering students

Mathematics 256 a course in differential equations for engineering students Mathematcs 56 a course n dfferental equatons for engneerng students Chapter 5. More effcent methods of numercal soluton Euler s method s qute neffcent. Because the error s essentally proportonal to the

More information

Use of Genetic Algorithms in Efficient Scheduling for Multi Service Classes

Use of Genetic Algorithms in Efficient Scheduling for Multi Service Classes Use of Genetc Algorthms n Effcent Schedulng for Mult Servce Classes Shyamale Thlakawardana and Rahm Tafazoll Centre for Communcatons Systems Research (CCSR), Unversty of Surrey, Guldford, GU27XH, UK Abstract

More information

A Semi-Distributed Load Balancing Architecture and Algorithm for Heterogeneous Wireless Networks

A Semi-Distributed Load Balancing Architecture and Algorithm for Heterogeneous Wireless Networks A Sem-Dstrbuted oad Balancng Archtecture and Algorthm for Heterogeneous reless Networks Md. Golam Rabul Ala Choong Seon Hong * Kyung Hee Unversty, Korea rob@networkng.khu.ac.kr, cshong@khu.ac.kr Abstract

More information

The Codesign Challenge

The Codesign Challenge ECE 4530 Codesgn Challenge Fall 2007 Hardware/Software Codesgn The Codesgn Challenge Objectves In the codesgn challenge, your task s to accelerate a gven software reference mplementaton as fast as possble.

More information

An Entropy-Based Approach to Integrated Information Needs Assessment

An Entropy-Based Approach to Integrated Information Needs Assessment Dstrbuton Statement A: Approved for publc release; dstrbuton s unlmted. An Entropy-Based Approach to ntegrated nformaton Needs Assessment June 8, 2004 Wllam J. Farrell Lockheed Martn Advanced Technology

More information

FAHP and Modified GRA Based Network Selection in Heterogeneous Wireless Networks

FAHP and Modified GRA Based Network Selection in Heterogeneous Wireless Networks 2017 2nd Internatonal Semnar on Appled Physcs, Optoelectroncs and Photoncs (APOP 2017) ISBN: 978-1-60595-522-3 FAHP and Modfed GRA Based Network Selecton n Heterogeneous Wreless Networks Xaohan DU, Zhqng

More information

Compiler Design. Spring Register Allocation. Sample Exercises and Solutions. Prof. Pedro C. Diniz

Compiler Design. Spring Register Allocation. Sample Exercises and Solutions. Prof. Pedro C. Diniz Compler Desgn Sprng 2014 Regster Allocaton Sample Exercses and Solutons Prof. Pedro C. Dnz USC / Informaton Scences Insttute 4676 Admralty Way, Sute 1001 Marna del Rey, Calforna 90292 pedro@s.edu Regster

More information

Improvement of Spatial Resolution Using BlockMatching Based Motion Estimation and Frame. Integration

Improvement of Spatial Resolution Using BlockMatching Based Motion Estimation and Frame. Integration Improvement of Spatal Resoluton Usng BlockMatchng Based Moton Estmaton and Frame Integraton Danya Suga and Takayuk Hamamoto Graduate School of Engneerng, Tokyo Unversty of Scence, 6-3-1, Nuku, Katsuska-ku,

More information

Load-Balanced Anycast Routing

Load-Balanced Anycast Routing Load-Balanced Anycast Routng Chng-Yu Ln, Jung-Hua Lo, and Sy-Yen Kuo Department of Electrcal Engneerng atonal Tawan Unversty, Tape, Tawan sykuo@cc.ee.ntu.edu.tw Abstract For fault-tolerance and load-balance

More information

Repeater Insertion for Two-Terminal Nets in Three-Dimensional Integrated Circuits

Repeater Insertion for Two-Terminal Nets in Three-Dimensional Integrated Circuits Repeater Inserton for Two-Termnal Nets n Three-Dmensonal Integrated Crcuts Hu Xu, Vasls F. Pavlds, and Govann De Mchel LSI - EPFL, CH-5, Swtzerland, {hu.xu,vasleos.pavlds,govann.demchel}@epfl.ch Abstract.

More information

Parallel matrix-vector multiplication

Parallel matrix-vector multiplication Appendx A Parallel matrx-vector multplcaton The reduced transton matrx of the three-dmensonal cage model for gel electrophoress, descrbed n secton 3.2, becomes excessvely large for polymer lengths more

More information

Determining the Optimal Bandwidth Based on Multi-criterion Fusion

Determining the Optimal Bandwidth Based on Multi-criterion Fusion Proceedngs of 01 4th Internatonal Conference on Machne Learnng and Computng IPCSIT vol. 5 (01) (01) IACSIT Press, Sngapore Determnng the Optmal Bandwdth Based on Mult-crteron Fuson Ha-L Lang 1+, Xan-Mn

More information

Scheduling and queue management. DigiComm II

Scheduling and queue management. DigiComm II Schedulng and queue management Tradtonal queung behavour n routers Data transfer: datagrams: ndvdual packets no recognton of flows connectonless: no sgnallng Forwardng: based on per-datagram forwardng

More information

CS 268: Lecture 8 Router Support for Congestion Control

CS 268: Lecture 8 Router Support for Congestion Control CS 268: Lecture 8 Router Support for Congeston Control Ion Stoca Computer Scence Dvson Department of Electrcal Engneerng and Computer Scences Unversty of Calforna, Berkeley Berkeley, CA 9472-1776 Router

More information

Research Article Adaptive Cost-Based Task Scheduling in Cloud Environment

Research Article Adaptive Cost-Based Task Scheduling in Cloud Environment Scentfc Programmng Volume 2016, Artcle ID 8239239, 9 pages http://dx.do.org/10.1155/2016/8239239 Research Artcle Adaptve Cost-Based Task Schedulng n Cloud Envronment Mohammed A. S. Mosleh, 1 G. Radhaman,

More information

The Greedy Method. Outline and Reading. Change Money Problem. Greedy Algorithms. Applications of the Greedy Strategy. The Greedy Method Technique

The Greedy Method. Outline and Reading. Change Money Problem. Greedy Algorithms. Applications of the Greedy Strategy. The Greedy Method Technique //00 :0 AM Outlne and Readng The Greedy Method The Greedy Method Technque (secton.) Fractonal Knapsack Problem (secton..) Task Schedulng (secton..) Mnmum Spannng Trees (secton.) Change Money Problem Greedy

More information

A QoS-aware Scheduling Scheme for Software-Defined Storage Oriented iscsi Target

A QoS-aware Scheduling Scheme for Software-Defined Storage Oriented iscsi Target A QoS-aware Schedulng Scheme for Software-Defned Storage Orented SCSI Target Xanghu Meng 1,2, Xuewen Zeng 1, Xao Chen 1, Xaozhou Ye 1,* 1 Natonal Network New Meda Engneerng Research Center, Insttute of

More information

Reliability and Energy-aware Cache Reconfiguration for Embedded Systems

Reliability and Energy-aware Cache Reconfiguration for Embedded Systems Relablty and Energy-aware Cache Reconfguraton for Embedded Systems Yuanwen Huang and Prabhat Mshra Department of Computer and Informaton Scence and Engneerng Unversty of Florda, Ganesvlle FL 326-62, USA

More information

A Binarization Algorithm specialized on Document Images and Photos

A Binarization Algorithm specialized on Document Images and Photos A Bnarzaton Algorthm specalzed on Document mages and Photos Ergna Kavalleratou Dept. of nformaton and Communcaton Systems Engneerng Unversty of the Aegean kavalleratou@aegean.gr Abstract n ths paper, a

More information

Delay Variation Optimized Traffic Allocation Based on Network Calculus for Multi-path Routing in Wireless Mesh Networks

Delay Variation Optimized Traffic Allocation Based on Network Calculus for Multi-path Routing in Wireless Mesh Networks Appl. Math. Inf. Sc. 7, No. 2L, 467-474 2013) 467 Appled Mathematcs & Informaton Scences An Internatonal Journal http://dx.do.org/10.12785/ams/072l13 Delay Varaton Optmzed Traffc Allocaton Based on Network

More information

ELEC 377 Operating Systems. Week 6 Class 3

ELEC 377 Operating Systems. Week 6 Class 3 ELEC 377 Operatng Systems Week 6 Class 3 Last Class Memory Management Memory Pagng Pagng Structure ELEC 377 Operatng Systems Today Pagng Szes Vrtual Memory Concept Demand Pagng ELEC 377 Operatng Systems

More information

Priority-Based Scheduling Algorithm for Downlink Traffics in IEEE Networks

Priority-Based Scheduling Algorithm for Downlink Traffics in IEEE Networks Prorty-Based Schedulng Algorthm for Downlnk Traffcs n IEEE 80.6 Networks Ja-Mng Lang, Jen-Jee Chen, You-Chun Wang, Yu-Chee Tseng, and Bao-Shuh P. Ln Department of Computer Scence Natonal Chao-Tung Unversty,

More information

Application of Improved Fish Swarm Algorithm in Cloud Computing Resource Scheduling

Application of Improved Fish Swarm Algorithm in Cloud Computing Resource Scheduling , pp.40-45 http://dx.do.org/10.14257/astl.2017.143.08 Applcaton of Improved Fsh Swarm Algorthm n Cloud Computng Resource Schedulng Yu Lu, Fangtao Lu School of Informaton Engneerng, Chongqng Vocatonal Insttute

More information

TECHNIQUE OF FORMATION HOMOGENEOUS SAMPLE SAME OBJECTS. Muradaliyev A.Z.

TECHNIQUE OF FORMATION HOMOGENEOUS SAMPLE SAME OBJECTS. Muradaliyev A.Z. TECHNIQUE OF FORMATION HOMOGENEOUS SAMPLE SAME OBJECTS Muradalyev AZ Azerbajan Scentfc-Research and Desgn-Prospectng Insttute of Energetc AZ1012, Ave HZardab-94 E-mal:aydn_murad@yahoocom Importance of

More information

Cluster Analysis of Electrical Behavior

Cluster Analysis of Electrical Behavior Journal of Computer and Communcatons, 205, 3, 88-93 Publshed Onlne May 205 n ScRes. http://www.scrp.org/ournal/cc http://dx.do.org/0.4236/cc.205.350 Cluster Analyss of Electrcal Behavor Ln Lu Ln Lu, School

More information

Module Management Tool in Software Development Organizations

Module Management Tool in Software Development Organizations Journal of Computer Scence (5): 8-, 7 ISSN 59-66 7 Scence Publcatons Management Tool n Software Development Organzatons Ahmad A. Al-Rababah and Mohammad A. Al-Rababah Faculty of IT, Al-Ahlyyah Amman Unversty,

More information

An Iterative Solution Approach to Process Plant Layout using Mixed Integer Optimisation

An Iterative Solution Approach to Process Plant Layout using Mixed Integer Optimisation 17 th European Symposum on Computer Aded Process Engneerng ESCAPE17 V. Plesu and P.S. Agach (Edtors) 2007 Elsever B.V. All rghts reserved. 1 An Iteratve Soluton Approach to Process Plant Layout usng Mxed

More information

Wishing you all a Total Quality New Year!

Wishing you all a Total Quality New Year! Total Qualty Management and Sx Sgma Post Graduate Program 214-15 Sesson 4 Vnay Kumar Kalakband Assstant Professor Operatons & Systems Area 1 Wshng you all a Total Qualty New Year! Hope you acheve Sx sgma

More information

Pricing Network Resources for Adaptive Applications in a Differentiated Services Network

Pricing Network Resources for Adaptive Applications in a Differentiated Services Network IEEE INFOCOM Prcng Network Resources for Adaptve Applcatons n a Dfferentated Servces Network Xn Wang and Hennng Schulzrnne Columba Unversty Emal: {xnwang, schulzrnne}@cs.columba.edu Abstract The Dfferentated

More information

Research Article. ISSN (Print) s k and. d k rate of k -th flow, source node and

Research Article. ISSN (Print) s k and. d k rate of k -th flow, source node and Scholars Journal of Engneerng and Technology (SJET) Sch. J. Eng. Tech., 2015; 3(4A):343-350 Scholars Academc and Scentfc Publsher (An Internatonal Publsher for Academc and Scentfc Resources) www.saspublsher.com

More information

Queueing Network-based Optimisation Techniques for Workload Allocation in Clusters of Computers *

Queueing Network-based Optimisation Techniques for Workload Allocation in Clusters of Computers * Queueng Network-based Optmsaton Technques for Workload Allocaton n Clusters of Computers * Lgang He, Stephen A. Jarvs, Davd Bacgalupo, Danel P. Spooner, Xnuo Chen and Graham R. Nudd Department of Computer

More information

Meta-heuristics for Multidimensional Knapsack Problems

Meta-heuristics for Multidimensional Knapsack Problems 2012 4th Internatonal Conference on Computer Research and Development IPCSIT vol.39 (2012) (2012) IACSIT Press, Sngapore Meta-heurstcs for Multdmensonal Knapsack Problems Zhbao Man + Computer Scence Department,

More information

Adaptive Energy and Location Aware Routing in Wireless Sensor Network

Adaptive Energy and Location Aware Routing in Wireless Sensor Network Adaptve Energy and Locaton Aware Routng n Wreless Sensor Network Hong Fu 1,1, Xaomng Wang 1, Yngshu L 1 Department of Computer Scence, Shaanx Normal Unversty, X an, Chna, 71006 fuhong433@gmal.com {wangxmsnnu@hotmal.cn}

More information

An Efficient Garbage Collection for Flash Memory-Based Virtual Memory Systems

An Efficient Garbage Collection for Flash Memory-Based Virtual Memory Systems S. J and D. Shn: An Effcent Garbage Collecton for Flash Memory-Based Vrtual Memory Systems 2355 An Effcent Garbage Collecton for Flash Memory-Based Vrtual Memory Systems Seunggu J and Dongkun Shn, Member,

More information

Real-time Fault-tolerant Scheduling Algorithm for Distributed Computing Systems

Real-time Fault-tolerant Scheduling Algorithm for Distributed Computing Systems Real-tme Fault-tolerant Schedulng Algorthm for Dstrbuted Computng Systems Yun Lng, Y Ouyang College of Computer Scence and Informaton Engneerng Zheang Gongshang Unversty Postal code: 310018 P.R.CHINA {ylng,

More information

3. CR parameters and Multi-Objective Fitness Function

3. CR parameters and Multi-Objective Fitness Function 3 CR parameters and Mult-objectve Ftness Functon 41 3. CR parameters and Mult-Objectve Ftness Functon 3.1. Introducton Cogntve rados dynamcally confgure the wreless communcaton system, whch takes beneft

More information

An Application of the Dulmage-Mendelsohn Decomposition to Sparse Null Space Bases of Full Row Rank Matrices

An Application of the Dulmage-Mendelsohn Decomposition to Sparse Null Space Bases of Full Row Rank Matrices Internatonal Mathematcal Forum, Vol 7, 2012, no 52, 2549-2554 An Applcaton of the Dulmage-Mendelsohn Decomposton to Sparse Null Space Bases of Full Row Rank Matrces Mostafa Khorramzadeh Department of Mathematcal

More information

Multi-objective Virtual Machine Placement for Load Balancing

Multi-objective Virtual Machine Placement for Load Balancing Mult-obectve Vrtual Machne Placement for Load Balancng Feng FANG and Bn-Bn Qu,a School of Computer Scence & Technology, Huazhong Unversty Of Scence And Technology, Wuhan, Chna Abstract. The vrtual machne

More information

Game Based Virtual Bandwidth Allocation for Virtual Networks in Data Centers

Game Based Virtual Bandwidth Allocation for Virtual Networks in Data Centers Avaable onlne at www.scencedrect.com Proceda Engneerng 23 (20) 780 785 Power Electroncs and Engneerng Applcaton, 20 Game Based Vrtual Bandwdth Allocaton for Vrtual Networks n Data Centers Cu-rong Wang,

More information

Improved Resource Allocation Algorithms for Practical Image Encoding in a Ubiquitous Computing Environment

Improved Resource Allocation Algorithms for Practical Image Encoding in a Ubiquitous Computing Environment JOURNAL OF COMPUTERS, VOL. 4, NO. 9, SEPTEMBER 2009 873 Improved Resource Allocaton Algorthms for Practcal Image Encodng n a Ubqutous Computng Envronment Manxong Dong, Long Zheng, Kaoru Ota, Song Guo School

More information

Configuration Management in Multi-Context Reconfigurable Systems for Simultaneous Performance and Power Optimizations*

Configuration Management in Multi-Context Reconfigurable Systems for Simultaneous Performance and Power Optimizations* Confguraton Management n Mult-Context Reconfgurable Systems for Smultaneous Performance and Power Optmzatons* Rafael Maestre, Mlagros Fernandez Departamento de Arqutectura de Computadores y Automátca Unversdad

More information

Internet Traffic Managers

Internet Traffic Managers Internet Traffc Managers Ibrahm Matta matta@cs.bu.edu www.cs.bu.edu/faculty/matta Computer Scence Department Boston Unversty Boston, MA 225 Jont work wth members of the WING group: Azer Bestavros, John

More information

Research of Dynamic Access to Cloud Database Based on Improved Pheromone Algorithm

Research of Dynamic Access to Cloud Database Based on Improved Pheromone Algorithm , pp.197-202 http://dx.do.org/10.14257/dta.2016.9.5.20 Research of Dynamc Access to Cloud Database Based on Improved Pheromone Algorthm Yongqang L 1 and Jn Pan 2 1 (Software Technology Vocatonal College,

More information

Avoiding congestion through dynamic load control

Avoiding congestion through dynamic load control Avodng congeston through dynamc load control Vasl Hnatyshn, Adarshpal S. Seth Department of Computer and Informaton Scences, Unversty of Delaware, Newark, DE 976 ABSTRACT The current best effort approach

More information

Efficient QoS Provisioning at the MAC Layer in Heterogeneous Wireless Sensor Networks

Efficient QoS Provisioning at the MAC Layer in Heterogeneous Wireless Sensor Networks Effcent QoS Provsonng at the MAC Layer n Heterogeneous Wreless Sensor Networks M.Soul a,, A.Bouabdallah a, A.E.Kamal b a UMR CNRS 7253 HeuDaSyC, Unversté de Technologe de Compègne, Compègne Cedex F-625,

More information

THere are increasing interests and use of mobile ad hoc

THere are increasing interests and use of mobile ad hoc 1 Adaptve Schedulng n MIMO-based Heterogeneous Ad hoc Networks Shan Chu, Xn Wang Member, IEEE, and Yuanyuan Yang Fellow, IEEE. Abstract The demands for data rate and transmsson relablty constantly ncrease

More information

Reducing Energy Consumption for Reconfiguration in Cloud Data Centers

Reducing Energy Consumption for Reconfiguration in Cloud Data Centers Reducng Energy Consumpton for Reconfguraton n Cloud Data Centers Invted Paper Omar Chakroun, Soumaya Cherkaou INTERLAB Research Laboratory, Unversté de Sherbrooke, Canada {omar.chakroun, soumaya.cherkaou}@usherbrooke.ca

More information

An Optimal Bandwidth Allocation and Data Droppage Scheme for Differentiated Services in a Wireless Network

An Optimal Bandwidth Allocation and Data Droppage Scheme for Differentiated Services in a Wireless Network Purdue Unversty Purdue e-pubs ECE Techncal Reports Electrcal and Computer Engneerng 3--7 An Optmal Bandwdth Allocaton and Data Droppage Scheme for Dfferentated Servces n a Wreless Network Waseem Shekh

More information

An Online Delay Efficient Multi-Class Packet Scheduler for Heterogeneous M2M Uplink Traffic

An Online Delay Efficient Multi-Class Packet Scheduler for Heterogeneous M2M Uplink Traffic An Onlne Delay Effcent Mult-Class Packet Scheduler for Heterogeneous M2M Uplnk Traffc Akshay Kumar, Ahmed Abdelhad and Charles Clancy Hume Center, Vrgna Tech Emal:{akshay2, aabdelhad, tcc}@vt.edu Abstract

More information

Hybrid Job Scheduling Mechanism Using a Backfill-based Multi-queue Strategy in Distributed Grid Computing

Hybrid Job Scheduling Mechanism Using a Backfill-based Multi-queue Strategy in Distributed Grid Computing IJCSNS Internatonal Journal of Computer Scence and Network Securty, VOL.12 No.9, September 2012 39 Hybrd Job Schedulng Mechansm Usng a Backfll-based Mult-queue Strategy n Dstrbuted Grd Computng Ken Park

More information

Achieving class-based QoS for transactional workloads

Achieving class-based QoS for transactional workloads Achevng class-based QoS for transactonal workloads Banca Schroeder Mor Harchol-Balter Carnege Mellon Unversty Department of Computer Scence Pttsburgh, PA USA @cs.cmu.edu Arun Iyengar Erch

More information

SRB: Shared Running Buffers in Proxy to Exploit Memory Locality of Multiple Streaming Media Sessions

SRB: Shared Running Buffers in Proxy to Exploit Memory Locality of Multiple Streaming Media Sessions SRB: Shared Runnng Buffers n Proxy to Explot Memory Localty of Multple Streamng Meda Sessons Songqng Chen,BoShen, Yong Yan, Sujoy Basu, and Xaodong Zhang Department of Computer Scence Moble and Meda System

More information

Dynamic Voltage Scaling of Supply and Body Bias Exploiting Software Runtime Distribution

Dynamic Voltage Scaling of Supply and Body Bias Exploiting Software Runtime Distribution Dynamc Voltage Scalng of Supply and Body Bas Explotng Software Runtme Dstrbuton Sungpack Hong EE Department Stanford Unversty Sungjoo Yoo, Byeong Bn, Kyu-Myung Cho, Soo-Kwan Eo Samsung Electroncs Taehwan

More information

Advanced radio access solutions for the new 5G requirements

Advanced radio access solutions for the new 5G requirements Advanced rado access solutons for the new 5G requrements Soumaya Hamouda Assocate Professor, Unversty of Carthage Tuns, Tunsa Soumaya.hamouda@supcom.tn IEEE Summt 5G n Future Afrca. May 3 th, 2017 Pretora,

More information

BioTechnology. An Indian Journal FULL PAPER. Trade Science Inc.

BioTechnology. An Indian Journal FULL PAPER. Trade Science Inc. [Type text] [Type text] [Type text] ISSN : 0974-74 Volume 0 Issue BoTechnology 04 An Indan Journal FULL PAPER BTAIJ 0() 04 [684-689] Revew on Chna s sports ndustry fnancng market based on market -orented

More information

Performance Comparison of a QoS Aware Routing Protocol for Wireless Sensor Networks

Performance Comparison of a QoS Aware Routing Protocol for Wireless Sensor Networks Communcatons and Network, 2016, 8, 45-55 Publshed Onlne February 2016 n ScRes. http://www.scrp.org/journal/cn http://dx.do.org/10.4236/cn.2016.81006 Performance Comparson of a QoS Aware Routng Protocol

More information

Adaptive Power-Aware Prefetch in Wireless Networks

Adaptive Power-Aware Prefetch in Wireless Networks Adaptve Power-Aware Prefetch n Wreless Networks Langzhong Yn, Student Member, IEEE, and Guohong Cao, Member, IEEE Abstract Most of the prefetch technques used n the current cache management schemes do

More information

Quantifying Responsiveness of TCP Aggregates by Using Direct Sequence Spread Spectrum CDMA and Its Application in Congestion Control

Quantifying Responsiveness of TCP Aggregates by Using Direct Sequence Spread Spectrum CDMA and Its Application in Congestion Control Quantfyng Responsveness of TCP Aggregates by Usng Drect Sequence Spread Spectrum CDMA and Its Applcaton n Congeston Control Mehd Kalantar Department of Electrcal and Computer Engneerng Unversty of Maryland,

More information

A MOVING MESH APPROACH FOR SIMULATION BUDGET ALLOCATION ON CONTINUOUS DOMAINS

A MOVING MESH APPROACH FOR SIMULATION BUDGET ALLOCATION ON CONTINUOUS DOMAINS Proceedngs of the Wnter Smulaton Conference M E Kuhl, N M Steger, F B Armstrong, and J A Jones, eds A MOVING MESH APPROACH FOR SIMULATION BUDGET ALLOCATION ON CONTINUOUS DOMAINS Mark W Brantley Chun-Hung

More information

WITH rapid improvements of wireless technologies,

WITH rapid improvements of wireless technologies, JOURNAL OF SYSTEMS ARCHITECTURE, SPECIAL ISSUE: HIGHLY-RELIABLE CPS, VOL. 00, NO. 0, MONTH 013 1 Adaptve GTS Allocaton n IEEE 80.15.4 for Real-Tme Wreless Sensor Networks Feng Xa, Ruonan Hao, Je L, Naxue

More information

Optimized Resource Scheduling Using Classification and Regression Tree and Modified Bacterial Foraging Optimization Algorithm

Optimized Resource Scheduling Using Classification and Regression Tree and Modified Bacterial Foraging Optimization Algorithm World Engneerng & Appled Scences Journal 7 (1): 10-17, 2016 ISSN 2079-2204 IDOSI Publcatons, 2016 DOI: 10.5829/dos.weasj.2016.7.1.22540 Optmzed Resource Schedulng Usng Classfcaton and Regresson Tree and

More information

A fair buffer allocation scheme

A fair buffer allocation scheme A far buffer allocaton scheme Juha Henanen and Kalev Klkk Telecom Fnland P.O. Box 228, SF-330 Tampere, Fnland E-mal: juha.henanen@tele.f Abstract An approprate servce for data traffc n ATM networks requres

More information

Tsinghua University at TAC 2009: Summarizing Multi-documents by Information Distance

Tsinghua University at TAC 2009: Summarizing Multi-documents by Information Distance Tsnghua Unversty at TAC 2009: Summarzng Mult-documents by Informaton Dstance Chong Long, Mnle Huang, Xaoyan Zhu State Key Laboratory of Intellgent Technology and Systems, Tsnghua Natonal Laboratory for

More information

Routing in Degree-constrained FSO Mesh Networks

Routing in Degree-constrained FSO Mesh Networks Internatonal Journal of Hybrd Informaton Technology Vol., No., Aprl, 009 Routng n Degree-constraned FSO Mesh Networks Zpng Hu, Pramode Verma, and James Sluss Jr. School of Electrcal & Computer Engneerng

More information

Design of the Application-Level Protocol for Synchronized Multimedia Sessions

Design of the Application-Level Protocol for Synchronized Multimedia Sessions Desgn of the Applcaton-Level Protocol for Synchronzed Multmeda Sessons Chun-Chuan Yang Multmeda and Communcatons Laboratory Department of Computer Scence and Informaton Engneerng Natonal Ch Nan Unversty,

More information

X- Chart Using ANOM Approach

X- Chart Using ANOM Approach ISSN 1684-8403 Journal of Statstcs Volume 17, 010, pp. 3-3 Abstract X- Chart Usng ANOM Approach Gullapall Chakravarth 1 and Chaluvad Venkateswara Rao Control lmts for ndvdual measurements (X) chart are

More information

Shared Running Buffer Based Proxy Caching of Streaming Sessions

Shared Running Buffer Based Proxy Caching of Streaming Sessions Shared Runnng Buffer Based Proxy Cachng of Streamng Sessons Songqng Chen, Bo Shen, Yong Yan, Sujoy Basu Moble and Meda Systems Laboratory HP Laboratores Palo Alto HPL-23-47 March th, 23* E-mal: sqchen@cs.wm.edu,

More information

FINDING IMPORTANT NODES IN SOCIAL NETWORKS BASED ON MODIFIED PAGERANK

FINDING IMPORTANT NODES IN SOCIAL NETWORKS BASED ON MODIFIED PAGERANK FINDING IMPORTANT NODES IN SOCIAL NETWORKS BASED ON MODIFIED PAGERANK L-qng Qu, Yong-quan Lang 2, Jng-Chen 3, 2 College of Informaton Scence and Technology, Shandong Unversty of Scence and Technology,

More information

Maintaining temporal validity of real-time data on non-continuously executing resources

Maintaining temporal validity of real-time data on non-continuously executing resources Mantanng temporal valdty of real-tme data on non-contnuously executng resources Tan Ba, Hong Lu and Juan Yang Hunan Insttute of Scence and Technology, College of Computer Scence, 44, Yueyang, Chna Wuhan

More information

Resource and Virtual Function Status Monitoring in Network Function Virtualization Environment

Resource and Virtual Function Status Monitoring in Network Function Virtualization Environment Journal of Physcs: Conference Seres PAPER OPEN ACCESS Resource and Vrtual Functon Status Montorng n Network Functon Vrtualzaton Envronment To cte ths artcle: MS Ha et al 2018 J. Phys.: Conf. Ser. 1087

More information

MobileGrid: Capacity-aware Topology Control in Mobile Ad Hoc Networks

MobileGrid: Capacity-aware Topology Control in Mobile Ad Hoc Networks MobleGrd: Capacty-aware Topology Control n Moble Ad Hoc Networks Jle Lu, Baochun L Department of Electrcal and Computer Engneerng Unversty of Toronto {jenne,bl}@eecg.toronto.edu Abstract Snce wreless moble

More information

An efficient iterative source routing algorithm

An efficient iterative source routing algorithm An effcent teratve source routng algorthm Gang Cheng Ye Tan Nrwan Ansar Advanced Networng Lab Department of Electrcal Computer Engneerng New Jersey Insttute of Technology Newar NJ 7 {gc yt Ansar}@ntedu

More information

Solutions for Real-Time Communication over Best-Effort Networks

Solutions for Real-Time Communication over Best-Effort Networks Solutons for Real-Tme Communcaton over Best-Effort Networks Anca Hangan, Ramona Marfevc, Gheorghe Sebestyen Techncal Unversty of Cluj-Napoca, Computer Scence Department {Anca.Hangan, Ramona.Marfevc, Gheorghe.Sebestyen}@cs.utcluj.ro

More information

A Fair Access Mechanism Based on TXOP in IEEE e Wireless Networks

A Fair Access Mechanism Based on TXOP in IEEE e Wireless Networks 11 Internatonal Journal of Communcaton Networks and Informaton Securty (IJCNIS) Vol. 8, No. 1, Aprl 16 A Far Access Mechansm Based on TXOP n IEEE 8.11e Wreless Networks Marjan Yazdan 1, Maryam Kamal, Neda

More information

High Performance DiffServ Mechanism for Routers and Switches: Packet Arrival Rate Based Queue Management for Class Based Scheduling

High Performance DiffServ Mechanism for Routers and Switches: Packet Arrival Rate Based Queue Management for Class Based Scheduling Hgh Performance DffServ Mechansm for Routers and Swtches: Packet Arrval Rate Based Queue Management for Class Based Schedulng Bartek Wydrowsk and Moshe Zukerman ARC Specal Research Centre for Ultra-Broadband

More information

DESIGNING TRANSMISSION SCHEDULES FOR WIRELESS AD HOC NETWORKS TO MAXIMIZE NETWORK THROUGHPUT

DESIGNING TRANSMISSION SCHEDULES FOR WIRELESS AD HOC NETWORKS TO MAXIMIZE NETWORK THROUGHPUT DESIGNING TRANSMISSION SCHEDULES FOR WIRELESS AD HOC NETWORKS TO MAXIMIZE NETWORK THROUGHPUT Bran J. Wolf, Joseph L. Hammond, and Harlan B. Russell Dept. of Electrcal and Computer Engneerng, Clemson Unversty,

More information

Virtual Machine Placement Based on the VM Performance Models in Cloud

Virtual Machine Placement Based on the VM Performance Models in Cloud Vrtual Machne Placement Based on the VM Performance Models n Cloud Hu Zhao, Qnghua Zheng, Member, IEEE, Wezhan Zhang Member, IEEE, Yuxuan Chen, Yunhu Huang SPKLSTN Lab, Department of Computer Scence and

More information

A Frame Packing Mechanism Using PDO Communication Service within CANopen

A Frame Packing Mechanism Using PDO Communication Service within CANopen 28 A Frame Packng Mechansm Usng PDO Communcaton Servce wthn CANopen Mnkoo Kang and Kejn Park Dvson of Industral & Informaton Systems Engneerng, Ajou Unversty, Suwon, Gyeongg-do, South Korea Summary The

More information

Assembler. Building a Modern Computer From First Principles.

Assembler. Building a Modern Computer From First Principles. Assembler Buldng a Modern Computer From Frst Prncples www.nand2tetrs.org Elements of Computng Systems, Nsan & Schocken, MIT Press, www.nand2tetrs.org, Chapter 6: Assembler slde Where we are at: Human Thought

More information

Decision Support for the Dynamic Reconfiguration of Machine Layout and Part Routing in Cellular Manufacturing

Decision Support for the Dynamic Reconfiguration of Machine Layout and Part Routing in Cellular Manufacturing Decson Support for the Dynamc Reconfguraton of Machne Layout and Part Routng n Cellular Manufacturng Hao W. Ln and Tomohro Murata Abstract A mathematcal based approach s presented to evaluate the dynamc

More information