Adaptive VM Management with Two Phase Power Consumption Cost Models in Cloud Datacenter

Size: px
Start display at page:

Download "Adaptive VM Management with Two Phase Power Consumption Cost Models in Cloud Datacenter"

Transcription

1 Mobile New Appl (2016) 21: DOI /s z Adapive VM Managemen wih Two Phase Power Consumpion Cos Models in Cloud Daacener Dong-Ki Kang 1 & Fawaz Al-Hazemi 1 & Seong-Hwan Kim 1 & Min Chen 2 & Limei Peng 3 & Chan-Hyun Youn 1 Published online: 31 March 2016 # Springer Science+Business Media New York 2016 Absrac As cloud compuing models have evolved from clusers o large-scale daa ceners, reducing he energy consumpion, which is a large par of he overall operaing expense of daa ceners, has received much aenion laely. From a cluser-level viewpoin, he mos popular mehod for an energy efficien cloud is Dynamic Righ Sizing (DRS), which urns off idle servers ha do no have any virual resources running. To maximize he energy efficiency wih DRS, one of he primary adapive resource managemen sraegies is a Virual Machine (VM) migraion which consolidaes VM insances ino as few servers as possible. In his paper, we propose a Two Phase based Adapive Resource Managemen (TP-ARM) scheme ha migraes VM insances from underuilized servers ha are supposed o be urned off o * Chan-Hyun Youn chyoun@kais.ac.kr Dong-Ki Kang dkkang@kais.ac.kr Fawaz Al-Hazemi fawaz@kais.ac.kr Seong-Hwan Kim s.h_kim@kais.ac.kr Min Chen Minchen2012@hus.edu.cn Limei Peng auroraplm@ajou.ac.kr School of Elecrical Engineering, KAIST, Daejeon, Souh Korea School of Compuer Science and Technology, Huazhong Universiy of Science and Technology, Wuhan , China Deparmen of Indusrial Engineering, Ajou Universiy, Suwon, Souh Korea susainable ones based on heir moniored resource uilizaions in real ime. In addiion, we designed a Self-Adjusing Workload Predicion (SAWP) mehod o improve he forecasing accuracy of resource uilizaion even under irregular demand paerns. From he experimenal resuls using real cloud servers, we show ha our proposed schemes provide he superior performance of energy consumpion, resource uilizaion and job compleion ime over exising resource allocaion schemes. Keywords Cloud compuing. Virual machine migraion. Dynamic righ sizing. Energy saving 1 Inroducion In modern cloud daa ceners, resource allocaion wih high energy efficiency has been a key problem because he operaing cos from energy consumpion has increased in recen years. According o an esimaion from [1], he cos of power and cooling has increased 400 % over 10 years, and 59 % of daa ceners idenify hem as key facors limiing server deploymens. Consequenially, challenges from energy consumpion and cooling have led o a growing push o achieve an energy efficien design for daa ceners. A he daa cener level, a promising echnology called Dynamic Righ Sizing (DRS) o save energy dynamically adjuss he number of acive servers (i.e., servers whose power is swiched on) in proporion o he measured user demands [2]. In DRS, energy saving can be achieved by enabling idle servers ha do no have any running VM insances o go ino low-power mode (i.e., sleep or shu down). In order o maximize he energy efficiency via DRS, one of he primary adapive resource managemen sraegies is VM consolidaion in which running VM insances can be dynamically inegraed ino he minimal

2 794 Mobile New Appl (2016) 21: number of cloud servers based on heir resource uilizaion colleced by a hypervisor monioring module [3]. Tha is, running VM insances on under-uilized servers, which are supposed o be urned off, could be migraed o powersusainable servers. However, i is difficul o efficienly manage cloud resources because cloud users ofen have heerogeneous resource demands underlying muliple service applicaions which experience highly variable workloads. Therefore, inconsiderae VM consolidaion using live migraion migh lead o undesirable performance degradaion due primarily o swiching overheads caused by he migraion and by urning off servers on [4]. Running service applicaions wih reckless VM migraion and DRS execuion could encouner serious execuion ime delays, increased laency or failures [5]. As a resul, a careful resource managemen scheme ha considers swiching overheads is necessary o reduce efficienly he energy consumpion of cloud servers while ensuring accepable Qualiy of Service (QoS) based on Service Level Agreemens o cloud users [6, 7]. In his paper, we propose a Two Phase based Adapive Resource Managemen (TP-ARM) scheme o address he above challenges for cloud daaceners. The energy consumpion model based on TP-ARM was formulaed wih a performance cos (repuaion loss) caused by an increased delay from downsizing acive servers, by an energy cosfrom keeping paricular servers acive, and by a cos incurred from swiching off servers on. Subsequenly, we designed an auomaed cloud resource managemen sysem called he Adapive Cloud Resource Broker (ACRB) sysem wih he TP-ARM scheme. Moreover, we inroduced our novel predicion mehod called Self-Adjusing Workload Predicion (SAWP) o increase he predicion accuracy of users fuure demands even under unsaic and irregular workload paerns. The proposed SAWP mehod adapively scales he hisory window size up or down according o he exraced workload s auocorrelaions and sample enropies which measure he periodiciy and bursiness of he workloads [8]. To invesigae he performance characerisics of he proposed approaches, we conduced various experimens o evaluae energy consumpion, resource uilizaion and compleion ime delay by live migraion and DRS execuion on a real esbed based on Opensack which is a well-known cloud plaform using KVM hypervisor [9]. Through meaningful experimenal resuls, we found ha our proposed TP-ARM scheme and SAWP mehod provide significan energy savings while guaraneeing accepable performance required by users in pracice. 2 Proposed adapive cloud resource brokering sysem In his secion, we discuss he archiecure of he auomaed Adapive Cloud Resource Brokering (ACRB) sysem. Our considered cloud environmen including ACRB, which suppors deploying a resource managemen scheme in order o migrae VM requess and adjuss he number of acive servers according o he workload level, is depiced in Fig. 1. There are m physical hoss and n VM requess in he cloud daacener. In he cloud daacener, he informaion on resource uilizaion is colleced hrough KVM hypervisor based monioring modules ino he Resource Monioring DB module, and repored o he ACRB which is responsible for solving he migraion of allocaed VM requess and sizing he daacener. The ACRB has wo modules: he TP-ARM module and he SAWP Manager. The TP-ARM module includes he VM Migraion Manager and he DRS Manager. In he firs phase, he VM Migraion Manager is responsible for selecing he appropriae VM requess o be migraed based on he measured resource uilizaion level. We describe he merics for deermining he VM reques migraion in deail in secion 4.In he second phase, he DRS Manager is responsible for finding he opimal number of acive servers in he cloud daacener. The DRS plan derived by he DRS Manager based on he amoun of submied VM requess and he measured resource uilizaion from each VM reques is delivered o he Daacener Manager, and he deermined percenage of idle servers are powered off. The SAWP Manager is responsible for adjusing he window size of hisorical daa adapively in order o predic he fuure demand of VM requess. The SAWP Manager is able o achieve he exac predicion of fuure demands even under varied workload levels by considering he periodiciy and he flucuaion of he hisorical daa. The owner of he cloud daacener has o minimize he coss for resource operaions while boosing he benefis which can increase based on he good repuaion of he observed QoS of he cloud services. In his paper, our ACRB ries o find a resource managemen soluion o minimize he oal cos of resource operaions including wo sub cos models: he energy consumpion cos and he performance repuaion cos. From he perspecive of he energy consumpion cos, he VM Migraion Manager ries o maximize he resource uilizaion of each physical hos by consolidaing running VM requess based on he whole offered load measured hrough he Libvir VM Monior module aached o each hos. The Libvir VM Monior module gauges he uilizaion of resources such as CPU, memory, and I/ O bandwidh in order o check wheher whole hoss are overloaded [10, 11]. VM requess on overloaded hoss are preferably migraed o oher hoss which have enough exra resource capaciy o accommodae newly allocaed VM requess. The DRS Manager sends he shudown messages o servers which have o be powered off, while i sends magic packes o he ones which have o be powered on again. Obviously, he ransiion of servers from sleeping mode o acive mode (awake ransiion) requires addiional energy consumpion (noe ha he ransiion overhead from acive o sleep (asleep ransiion)

3 Mobile New Appl (2016) 21: Fig. 1 Adapive Cloud Resource Brokering (ACRB) sysem including TP-ARM module and SAWP manager for green cloud daaceners can be negligible because i requires only a shor ime o be carried ou compared o he awake ransiion). Therefore, i is clear ha frequen awake ransiions have o be discouraged in order o minimize unnecessary energy consumpion. To simplify our model, we assume ha he asleep ransiion causes no addiional energy consumpion (in fac, i requires non-zero energy consumpion). In erms of he performance repuaion cos, he VM Migraion Manager ries o decenralize running VM requess over muliple physical hoss in order o avoid QoS deerioraion caused by VM inerference. In cloud daaceners, he VM co-locaion inerference is he key facor ha makes servers undergo severe performance degradaion [12, 13]. VM co-locaion inerference is caused by resource conenion which is refleced mainly by he number of co-locaed VM insances and he resource uilizaion of hem. Briefly, VM co-locaion inerference becomes larger as more VM insances are co-locaed on a common server, and subsequenly, higher resource uilizaion occurs. Therefore, VM requess have o be scaered in order o avoid performance degradaion by VM co-locaion inerference as bes as possible. Because of he complexiy of he opimizaion for resource managemen in large-scale cloud daaceners, our TP-ARM scheme adops a meaheurisic based on GA o obain a near opimal soluion for VM migraion o achieve energy savings and QoS assurance in a cloud daacener. 3 Proposed model for cos opimizaion of TP-ARM in he cloud In his secion, we inroduce an energy cos model and performance repuaion cos model based on our proposed TP-ARM scheme including VM live migraion and DRS execuion in he ACRM. 3.1 Workload cos model for phase 1: VM migraion In general, VM requess have heerogeneous workloads in cloud daaceners. There are hree ypes of VM reques workloads: CPU, block I/O, and nework I/O inensive workloads. The CPU inensive workloads such as scienific applicaions, high-definiion video compression or big daa analysis should be processed wihin an accepable compleion ime deermined by he cloud service users. The block I/O inensive workloads such as huge engineering simulaions for criical areas include asrophysics, climae, and high energy physics which are highly daa inensive [14]. These applicaions conain a large number of I/O accesses where large amouns of daa are sored o and rerieved from disks. Nework I/O inensive workloads such as Inerne web services and mulimedia sreaming services have o be processed wihin a desirable response ime according o heir applicaion ype [10, 15]. Each

4 796 Mobile New Appl (2016) 21: VM reques requires differen resource uilizaions according o heir running applicaions. We caegorized he resource uilizaion of running VM requess on a physical hos in a cloud daacener ino wo pars in his paper, he Flavor Uilizaion (FU) and he Virual Uilizaion (VU). FU represens he raio of he resource flavor (i.e., he specificaion of he resource requiremen) of a VM reques o he resource capaciy of he physical hos. VU represens he resource uilizaion of an k assigned virual resource. The Flavor Uilizaion FU j, i of a VM reques i on a physical hos j in he resource componen k k and he Virual Uilizaion VU j, i of he VM reques i on he physical hos j in he resource componen k are given by FU k j; i ¼ flvk i rcp k j RU k j; i ¼ FUk j; i VUk j; i ð1þ ð2þ where flv k i is he flavor of he VM reques i in he resource componen k; rcp k j is he resource capaciy of he physical hos k j in he resource componen k,andru j, i is he acual resource uilizaion of he VM reques i on he physical hos j in he resource componen k; herefore, i describes he pracical workload in he resource componen k. Noe ha he FU of he VM reques can be deermined hrough is aached service descripion o he ACRM in advance, while he VU can only be measured by he inernal monioring module in he cloud server during he running duraion of he VM reques. In he period, he VM migraion plan is designed according o he FU of he submied VM requess a period and he measured RU of each hos in he pas hisory 1. The VM migraion plan should saisfy he following consrains: flv k i S kðþ; i i; k; ð3þ flv k i 0; i; k ð4þ S saeðþ¼1; i i; ð5þ where S k (i) denoes he remain capaciy of resource k in he physical server S which is assigned o he VM reques i a ime period,ands sae (i) represens he sae of he physical server S which is assigned o he VM reques i a ime period. Ifhe server S is in he sleep mode, hen S sae (i) = 0, oherwise, S sae (i) = 1 (i.e., i is in he acive mode). Consrain (3) represens ha he oal requiremens of he resource capaciy of he newly allocaed VM requess and migraed VM requess a ime period canno exceed he resource capaciy provided by heir assigned physical server S. Nex, we consider a VM performance repuaion which is deermined based on he RU of each physical server. The VM co-locaion inerference implies ha he virualizaion of he cloud suppors resource isolaion explicily when muliple VM requess are running simulaneously on a common PM however, i does no mean he assurance of performance isolaion beween VM requess inernally. There is a srong relaionship beween VM co-locaion inerference and boh of he number of co-locaed VMs and heir resource uilizaion in he PM. As he number of colocaed VM requess increase and he RU of he VM requess becomes larger, VM co-locaion inerference becomes more severe. The VM Migraion Manager selecs server candidaes o be migraed based on he amoun of RU for each physical server. To achieve his, he RU size of each server is measured hrough an inernal monioring module, and he VM Migraion Manager checks wheher hey exceed he predeermined RU hreshold value RU hr. In servers which have an RU size over he RU hr, hey are considered as candidaes for migraion servers during he nex period. Afer he migraion servers are chosen, he VM requess o be migraed and heir desinaion servers are deermined based on he performance repuaion cos model. We propose he objecive funcion of he performance repuaion cos C repu a ime C repu given by 0 1 X PM þ j jj C repu ¼ ρ X X inf ω k PM RU k i¼1 idx j j; i 1C A ð6þ j k þρ mig T mig X j PM 1 j RU k hr PM j where (x) + = max(x, 0), and ρ inf and ρ mig are he price consans of he VM inerference cos and VM migraion cos, respecively. We use T mig o denoe he processing ime for he VM migraion. The firs erm in he righ hand of Eq. (6) represens he following: as he number of concurren running VM requess on a physical server increases, he number of users experiencing undesirable performance degradaion also increases. The second erm represens he following: as he number of migraed VM requess increases, he migraion overhead is also increases. Therefore, a migraion plan needs o be found ha saisfies boh avoiding unnecessary migraion overhead and minimizing performance repuaion degradaion. 3.2 Energy consumpion cos model for phase 2: DRS procedure To achieve a power-proporional cloud daacener which consumes power only in proporion o he workload, we considered a DRS procedure which adjuss he number of acive servers by urning hem on or off dynamically [2]. Obviously, here is no need o urn all he servers in a cloud daacener on when he oal workload is low. In he DRS procedure, he sae of servers which have no running applicaions can be ransied o he power saving mode (e.g., sleep or hibernaion) in order o avoid wasing energy. A each ime, he number of acive servers in he cloud daacener is deermined by a DRS procedure plan according o he workload

5 Mobile New Appl (2016) 21: of he allocaed VM requess. In order o successfully deploy he DRS procedure ono our sysem, we considered he swiching overhead for adjusing he number of acive servers (i.e., for urning servers in sleep mode on again). The swiching overhead includes he following: 1) addiional energy consumpion from he ransiion from a sleep o acive sae (i.e., awaken ransiion); 2) wear-and-ear cos of he server; 3) faul occurrence by urning servers in sleep mode [2]. We considered he energy consumpion as he overhead from DRS execuion. Therefore, we define he consan P awake o denoe he amoun of energy consumpion for he awake ransiion of he servers. Then, he oal energy consumpion C energy of he cloud daacener a ime is given by C energy ¼ ρ X pp acive PM j j X X þ þ ρ p P awake T swich j PM j j PM j ð7þ where (x) =min(x, 1). We use ρ p o denoe he consan of he power consumpion price, and P acive and P awake are he amoun of power consumpion for he acive mode and he awake ransiion of he server, respecively. T swich is he ime requiremen for he awake ransiion. The firs erm on he righ side of Eq. (7) represens he energy consumpion for using servers o serve VM requess allocaed o all he physical servers in he cloud daacener a ime, and he second erm represens he energy consumpion for he awaken ransiion of sleeping servers. Especially, he second erm implies ha a frequen change in he number of acive servers could increase an undesirable wase of energy. Noe ha he overhead by he ransiion from he acive o he sleep sae (i.e., asleep ransiion) is ignored in our model because he ime required for he asleep ransiion is relaively shor compared o he one for he awaken ransiion. The purpose of our algorihm is o achieve a desirable VM migraion and DRS procedure plan ha are energy efficien as well as a QoS aware resource managemen a each period ieraively. A period 1, he proposed TP-ARM approach aims o minimize he cos funcion in Eq. (8) byfinding he soluion PM as follows, minimize PM C oal ¼ C repu þ C energy subjec o: Eq ð3þ; ðþand 4 ð5þ ð8þ To solve he objecive cos funcion in Eq. (8), we prefer a well-known evoluionary meaheurisic called he Geneic Algorihm (GA) in order o approximae he opimal plan P M a each period because Eqs. (6) and(7) havenon-linear characerisics. In nex secion, our proposed TP-ARM wih he VM migraion and DRS procedure is inroduced in deail. 4 Heurisic algorihms for he proposed TP-ARM scheme In his secion, we describe heurisic algorihms for he proposed TP-ARM scheme discussed in Algorihms 1, 2, and 3 shown in Figs. 2, 3, and4. Firs, he process of VM migraion in he TP- ARM approach includes he following four seps: 1) monior and collec he resource uilizaion daa on VM requess for each physical server hrough he aached Libvir based monioring ools; 2) go sep 3 if he average uilizaion of all he acive servers are significanly low (i.e., below he predefined hreshold), oherwise go o sep 4; 3) choose acive servers which are supposed o be urned off, migrae all he VM insances on hem o oher servers and rigger DRS execuion; 4) deermine he number of sleeping servers which are supposed o be urned on and send magic packes o hem for wake-up if he average uilizaion of all he acive servers are significanly high (i.e., above he predefined hreshold), oherwise mainain he curren number of acive servers. Algorihm 1 shows he procedure for Phase 1:VM migraion in he TP-ARM scheme. From line 00 o 06, he resource uilizaion RU of all he VM requess on he physical servers in he cloud daacener is measured by he Libvir API monioring module. The average resource uilizaion RU k a all resource componens k is calculaed in line 07. If he RU k is below he predeermined RU k a all he resource componens, hen he hr low opimal VM migraion plan PM is derived, and he algorihm is finished. Oherwise, if he RU k exceeds he predeermined RU k, hen Algorihm 2 for he DRS procedure is riggered. hr high Algorihm 2 shows he procedure for Phase 2: DRS in he TP-ARM scheme. In line 00, he hisory window size ϖ is deermined by he ACRM hrough Algorihm 4: he SAWP scheme ha is described in nex secion. In line 02, we calculae he sign φ and he angle ξ of he slope of he hisorical resource uilizaion curve from he curren ime 1oheime ϖ.if φ 0 (i.e., he resource uilizaion is increased), and hen, we ge he coefficien α based on α large ξ (α large > α small ) o adapively reac o he large workload level in he cloud daacener. Oherwise, if φ < 0 (i.e., he resource uilizaion is decreased), hen we ge α based on α small ξ o maximize he energy saving of he cloud daacener. In line 08, we deermine he number of servers o be in acive mode in he nex period. Algorihm 3 describes he GA for he TP-ARM scheme in deail. pop h is a populaion wih size P (even number) a he h h generaion. h max is he maximum of he GA ieraion coun. In line 03, f( ) is a finess funcion of he oal cos wih wo parameers, candidae soluion PM h; i and he previous soluion P M 1 a ime 1. From line 04 o 06, he wo candidae soluions are ieraively chosen randomly from pop o generae offsprings by crossover unil here are no remaining unseleced soluions in pop. From line 07 o 08, he finess funcion values of each offspring are calculaed similar o line 03. In line 09, all

6 798 Mobile New Appl (2016) 21: Algorihm 1. Phase 1: VM migraion in TP-ARM Inpu : he VM requess allocaion of each server M, Oupu : he VM migraion plan M, 00: for each, 01: for each 02: for each resource componen 03: measure resource uilizaion of he VM reques wih 04: end for 05: end for 06: end for 07: calculae he average resource uilizaion a all resource componens 08: if hen 09: derive he VM migraion plan based on Eq.(8) hrough Algorihm 3. 10: else if hen 11: go o Algorihm 2. 12: end if Fig. 2 Phase 1: VM migraion procedure consolidaes running VM insances from he low-uilized servers o ohers in order o maximize he energy saving of he cloud daacener he paren soluions in pop and generaed offsprings are sored in ascending order for heir corresponding finess funcion values. In line 10, he nex populaion including only P soluions ha achieve good performance from he union of he original pop and derived offsprings is generaed o improve he qualiy of he final soluion. From line 11 o 12, o reduce he ime complexiy of he GA procedures, when we encouner he firs soluion ha has a finess funcion value below he predeermined finess hreshold value fv hr, i couns as a final soluion for he nex period, and he algorihm is finished. In line 14, if we canno find a soluion ha saisfies fv hr when he ieraion coun reaches h max,henweselec a soluion ha has a minimum finess funcion value in he populaion which saisfies he condiions in Eqs. (3), (4) and(5) as a final soluion for he nex period. If here are no soluions o saisfy all he consrains, hen we jus preserve he curren resource allocaion vecor shown from line 15 o 16. In he populaion of a GA, muaions are ofen applied in order o include new characerisics from he offsprings ha are no inheried rais from he parens [16]. We did no consider muaions in our GA in his paper; however, i can be used o improve he qualiy of he GA for he TP-DRM in fuure work. 5 Self-adjusing workload predicion scheme Figure 5 describes he procedure of he proposed SAWP algorihm in ACRB. The irregulariy funcion g(ε, δ) represens a level of unpredicabiliy for fuure resource Algorihm 2. Phase 2: DRS procedure in TP-ARM Inpu : he average resource uilizaion he VM requess allocaion of each server he hisorical daa of resource uilizaion in Oupu : he se of powered-off servers o be urned on a ime. 00: deermine he boundary size of he hisory window, ( ) hrough Algorihm 4. 01: esimae he sign and angle ξ of an slope of he hisorical resource uilizaion curve from o. 02: if hen 03: α = α ξ 04: end if 05: if hen 06: α = α ξ 07: end if 08: deermine he number of sleeping servers supposed o be urned on according o 09: derive he se of powered-off servers o be urned on a ime Fig. 3 Phase 2: DRS procedure urns sleeping servers on in order o reac he increased workload level

7 Mobile New Appl (2016) 21: Algorihm 3. GA for searching VM migraion plan in TP-ARM Inpu : he se of VM requess allocaion of whole servers M a ime Oupu : he se of VM requess migraion plan, M a ime 00: iniialize =, M,, M,, and even number 01: while h h 02: for each 03: 04: while 05: selec parens randomly from 06: = crossover 07: for each 08: 09: sor in ascending order of soluions' corresponding 10: 11: if hen 12: and exi 13: h++ 14: =argmin (3), (4), (5) 15: if M = hen 16: Fig. 4 GA for VM migraion deermines he appropriae VM requess supposed o be migraed in order o maximize he energy saving of he cloud daacener uilizaion where ε is is flucuaion value (i.e., levels of insabiliy and aperiodiciy), and δ is is bursiness value (i.e., levels of sudden surge and decline), and boh values are calculaed based on [8]. high According o he predeermined hreshold values g hr and g low hr wih g(ε, δ), he hisory window size ϖ 'isadapively updaed a each predicion process. A relaively long hisory window size is no suiable o reac o recen changes in he workload bu is oleran o varied workload paerns over a shor ime while a shor hisory window size is favorable o efficienly respond o he laes workload paerns bu is no good for widely varying workloads. Consequenly, he SAWP algorihm generally ouperforms radiional predicion schemes during drasic uilizaion changes from various cloud applicaions because i is able o cope wih emporary resource uilizaions (i.e., does no reflec overall rends) by adjusing he hisory window size ϖ '. Algorihm 4. Self Adjusing Workload Predicion scheme Inpu : he hisorical daa of resource uilizaion in Oupu : he boundary size of hisory window 00: 01: analyze he hisorical daa of resource uilizaion from o 02: exrac he flucuaion size and he bursness value based on [6] 03: if ( ) hen 04: increase based on ( ) 05: end if 06: if ( ) hen 07: decrease based on ( ) 08: end if 09: predic he fuure demand based on from o Fig. 5 SAWP scheme is able o increase he accuracy of he predicion even under he dynamic workload

8 800 Mobile New Appl (2016) 21: Experimenal resuls and discussion In our experimens, we measured various merics which affec he parameer decision for our proposed algorihms. To his end, we esablished five cluser servers as a cloud plaform, one server for ACRB wih a Mysql DB sysem, a power measuring device from Yoco-Wa [17] and a lapop machine called he VirualHub for collecing and reporing informaion on he measured power consumpion shown in Fig. 6. The hardware specificaion of each server for he cloud compue hos is as follows: an Inel i (8-cores, 3.4Ghz), 16 GB RAM memory, and wo 1 Gbps NICs (Nework Inerface Cards). In order o measure efficienly he power consumpion of he cloud cluser server, we used a power measuring device model called YWATTMK1 by Yoco-Wa. This model has a measuremen uni of 0.2 W for AC power wih an error raio of 3 % and W for he DC power wih an error raio of 1.5 %. The VirualHub collecs informaion on power consumpion from YWATTMK1 hrough he Yoco-Wa Java API and repors i o he power monioring able of he Mysql DB sysem in he ACRB periodically. The dynamic resource uilizaions by each VM insance are measured via our developed VM monioring modules based on he Libvir API and are sen o he resource monioring able of he Mysql DB sysem periodically. In addiion, a SATA 3 TB hard disk called he G-drive was deployed as a NFS server in our esbed for live migraion [18]. We adoped Opensack kilo version which is a well-known open source soluion based on KVM Hypervisor as a cloud plaform in our esbed. Finally, we used Powerwake package [19] o urn remoely off servers on via Wake on Lan (WOL) echnology for DRS execuion. In Table 1, we show he average power consumpion and resource uilizaion of wo running applicaions: Monage m projecion and fp ransfer. Monage projec is an open source based scienific applicaion, and i has been invoked by NASA/IPAC Infrared Science Archive as a oolki for assembling Flexible Image Transpor Sysem (FITS) images ino cusom mosaics [20, 21]. The m projecion in Monage is a cpu-inensive applicaion while he fp ransfer is a nework-inensive one. Therefore, he running of m causes a power consumpion of abou 75 Wh and a cpu uilizaion of abou 15 % whereas he power consumpion by fp ransfer for a 1.5 GB es.avi file is abou 60 Wh, and he nework bandwidh usage is abou 3.7Mbps. Tha is, he cpu usage is he main par ha affecs he power consumpion of server. In erms of DRS execuion, he power consumpion by an off server is abou 2.5 Wh (noe ha his value is no zero because he NIC and some is peripheral componens are sill powered on o mainain he sandby mode o receive he magic packes from he Powerwake conroller) while he asleep and awaken ransiion procedures, which cause he swiching overhead for DRS, require a power consumpion of abou 80 Wh o urn he acive servers off or o urn off servers on, respecively. Fig. 6 Experimenal Environmen

9 Mobile New Appl (2016) 21: Table 1 Average power consumpion and resource uilizaion by Monage and fp ransfer Hos sae Power Consumpion (Wh) CPU uilizaion Mem uilizaion Ne bandwidh Idle 55 Wh 1.5 % 3 % 20 Kbps (byes) Acive Monage m (projecion) 75 Wh 15 % 3.7 % 20 Kbps (byes) es.avi downloading (fp, 1.5GB) 60 Wh 2 % 3.7 % 3.7 Mbps (byes) asleep (o Power Off) Wh (5 7 s) 1.8%(5 7 s) 3%(5 7 s) 20 Kbps (byes) Power Off (hibernaing) 2.5 Wh awake (from Power Off) 78 Wh (50s-1min) The asleep ransiion procedure is rivial because i requires a shor ime (i.e., 5 ~ 7 s) o complee even hough is power consumpion is considerable whereas he awaken ransiion procedure requires a relaively long execuion ime (i.e., more han 1 min) and should be considered carefully. The overhead for he awaken ransiion would be a more serious problem in pracice because is required execuion ime is generally far longer (i.e., above 10 min) for muliple servers of racks in a daacener. Therefore, i is essenial o consider he swiching overhead for awaken ransiion o reduce efficienly he resource usage cos of he daacener. Figure 7 shows he addiional energy consumpion overhead of a single physical server from he DRS execuion wih respec o he asleep ransiion (i.e., from he acive mode o he sleep mode) and he awaken ransiion (i.e., from he sleep mode o he acive mode). Noe ha he asleep ransiion requires a relaively shor processing ime of abou 7 or 8 s and causes an addiional energy consumpion of abou 20 % compared o he idle sae of he server, while he awaken ransiion needs 60 or 90 s o be carried ou. Alhough no included in his paper, we also verified ha some of he HP physical servers in our laboraory required more han 10 min o be urned on. We expec ha physical servers in a real cloud daacener require ens of minues or hundreds of minues o ge back from he sleep mode o he acive mode. Moreover, he awake ransiion causes an addiional energy consumpion of abou 40 % on average compared o he idle sae of a server. These resuls imply ha he frequen DRS execuion migh cause he degradaion of he energy saving performance in a cloud daacener. Fig. 7 Power Consumpion Overhead from DRS execuion of he es server wih respec o he asleep and awaken ransiion Our proposed TP-ARM scheme is able o avoid he unnecessary energy consumpion overhead by he awaken ransiion hrough he cos model by considering he DRS ransiion overhead shown in Eq. (7). Figure 8 shows he resource uilizaion and he power consumpion measured by he Libvir API monioring module and Yoco-Wa power measuring device. VM reques insance-10 runs m mproj, insance- 0f runs m , and insance-0c runs he sreaming server wih he 180 MB movie file. Figure 8 (a) ~ (c) shows ha heir resource uilizaion is consisen wih he resource inensive characerisic of heir running workload. Figure 8d shows he differen energy consumpions according o each workload ype. The energy consumpion of he physical server is 60 Wh in he idle sae; i is abou 82 Wh when running m and 95 Wh when running boh m and m , while he sreaming server jus causes an addiional energy consumpion of abou 2 Wh. As menioned in secion 3, he main par of he resource componens affecing he power consumpion in a physical server is he CPU resource, and he effecs of oher componens such as he memory, sorage, and nework inerface cards are negligible in general. These resuls are consisen wih he ones in Table 1. Therefore, hese resuls imply ha he energy consumpion can be differen according o which resource componen has high uilizaion. Our proposed workload cos model of he TP-ARM scheme considers differen weigh values for each resource componen in order o derive he pracical cos of he resource managemen. Figure 9 shows he performance of he CPU uilizaion for he VM live migraionbeween he source machine (kdkcluser2) and he desinaion machine (kdkcluser1). There are VM reques insances, such as running insance-33 and 34, which execue he compression of video files 5 GB and 4 GB in size, respecively. Insance-33 execued he process of m mproj a 12:40:14 local ime and sared migraion o he kdkcluser1 a 12:44:00 local ime. The migraion of insance-33 was compleed by 13:24:30 local ime, and is execuion of m mproj ended a 13:47:16. Namely, he compleion ime of m mproj a insance-33 was 67 min in oal, and is migraion imes were over 30 min. If we consider he ime o complee m mproj in he case of no VM live migraion, i is forecased o be around 10 min. We can see ha

10 802 Mobile New Appl (2016) 21: Fig. 8 Libir API monioring module shows he resuls of resource uilizaion of running VM requess in he kdkcluser1 in (a), (b) and(c). Yoco-Wa module [17] was used o measure he power consumpion of he kdkcluser1 as shown in (d) here exis quie big overheads in VM live migraion. In addiion, VM migraion shows some problems in power consumpion. We considered cos models o idenify an efficien migraion scheme whichwasdefinedineq.(6). Figure 10 shows he es environmens and he operaion of he VM migraion when he proposed TP-ARM schemes are applied o he es cloud sysems wih heerogeneous applicaions and es programsincluding m , pbzip2, and neperf. From he experimens, we obained ineresing resuls in he comparison of he uilizaion performance, which examined he CPU uilizaion monioring resuls and he measured power consumpion, respecively, shown in Fig. 11. From he resuls, we found rapidly changing insans of uilizaion when he cloud brokering sysem considered he power consumpions for each running applicaion under cloud daa cener environmens. Addiionally, he es se neperf, a nework inensive workload of a es cloud sysem (e.g., kdkcluser4 wih insance-35) in Fig. 11d, showed very low performance in CPU uilizaion. I was enough o saisfy he hreshold value o run he VM migraion efficienly. And hen, he proposed TP-ARM algorihm adjuss he VM migraion procedure o reduce he power consumpion sufficienly. Basically, he TP-ARM is riggered o migrae insance-35 o anoher sysem kdkcluser2; hereafer, i changes he kdkcluser s sleeping mode in advance. Nex, because insance-34 of he kdkcluser Fig. 9 Resuls of CPU uilizaion of he migraed VM reques from (a) a source server:kdkcluser2 o (b) a desinaion server:kdkcluser1

11 Mobile New Appl (2016) 21: Fig. 10 Tes environmens and operaion of he VM migraion when he proposed TP-ARM schemes are applied for es cloud servers (e.g., kdkcluser1 ~ 5 in lab es environmens) wih heerogeneous applicaions and es programs, such as m , pbzip2, and neperf. (a) Anexample of es schedule under cloud esbed environmens (b) Illusraion of he VM migraion using wo-phase power consumpion cos models (a) The schedule of he use case (b) The illusraion of he use case sysem was running neperf, which generaed a very low performance of uilizaion, he TP-ARM also did a migraion of he insance-34 o he kdkcluser2 and ransied o he sleeping mode as well. Therefore, we could expec efficien power consumpion hrough he change in he sleeping mode of kdkcluser3 and kdkcluser4, respecively, as well as keeping acive modes for boh kdkcluser1 and kdkcluser2. Figure 11 shows he performance comparison of he power consumpion for he VM migraion and DRS process based on he TP-ARM algorihm, e.g., seen in Fig. 10. We can see a performance improvemen of 60 Wh each in power consumpion afer he change in he sleeping mode in he case of he VM migraion requess. Through he experimens, we verified ha he proposed TP- ARM scheme, which carries ou adapive migraion and asleep ransiion hrough real-ime monioring of resource uilizaion, has good performance in reducing he power consumpion effecively. This sudy provides good resuls for achieving an energy efficien cloud service for users by no increasing QoS degradaion as much. 7 Conclusion In his paper, we inroduced an Adapive Cloud Resource Broker (ACRB) wih Two Phases based he Adapive Resource Managemen (TP-ARM) scheme for energy efficien resource managemen by real ime based VM monioring in a cloud daa cener. Our proposed approach is able o reduce efficienly he energy consumpion of he servers wihou a significan performance degradaion by live migraion and Dynamic Righ Sizing (DRS) execuion hrough a considerae model ha considers swiching overheads. The various experimenal resuls based on he Opensack plaform sugges ha our proposed algorihms can be deployed o prevalen cloud daa ceners. The novel predicion

12 804 Mobile New Appl (2016) 21: Fig. 11 Performance comparison of CPU uilizaion using 4 es sysems in lab environmens. (a), (b), and (c) show he resuls of he measured uilizaion during he VM live migraion in es use case as shown in Fig. 9. (d) shows uilizaion of he proposed TP-ARM scheme ha was used in es cloud environmens. (e) Measured power consumpion of es cloud sysems (e.g., kdkcluser1 ~ 4) in lab cloud es environmens mehod called Self Adjusing Workload Predicion (SAWP) is proposed in order o improve he accuracy of forecasing fuure demands even under drasic workload changes. Especially, we evaluaed he performance of our proposed TP-ARM scheme hrough various applicaions such as Monage, pbzip2, neperf, and he sreaming server which

13 Mobile New Appl (2016) 21: have heerogeneous workload demands. Our TP-ARM scheme could maximize he energy saving performance of he DRS procedure by Phase 1: VM migraion achieves consolidaed resource allocaion under a low workload level. Moreover, i could ensure he QoS of cloud service users by Phase 2: he DRS procedure increases adapively he number of acive servers under a high workload level. Through experimens based on a pracical use case, our proposed scheme is no only feasible from a heoreical poin of view bu also pracical in a real cloud environmen. In fuure work, we will demonsrae ha our proposed algorihm ouperforms exising approaches for energy efficien resource managemen hrough various experimens based on he implemened sysem in pracice. Acknowledgmens This work was suppored by BThe Cross-Minisry Giga KOREA Projec^ of he Minisry of Science, ICT and Fuure Planning, Korea [GK13P0100, Developmen of Tele-Experience Service SW Plaform based on Giga Media]. References 1. Inernaional Daa cener Corporaion, hp:// 2. Lin M, Wierman A, Andrew LLH, Thereska E (2013) Dynamic righ-sizing for power-proporional daa ceners. IEEE/ACM Trans Neworking 21(5): Xiao Z, Song W, Chen Q (2013) Dynamic resource allocaion using virual machines for cloud compuing environmen. IEEE Trans Parallel Disrib Sys 24(6): Beloglazov A, Buyya R (2012) Opimal online deerminisic algorihms and adapive heurisics for energy and performance efficien dynamic consolidaion of virual machines in cloud daa ceners. Concurr Compu: Prac Experience 24: doi: / cpe Kang DK, Hazemi FA, Kim SH, Youn CH (2015) Dynamic virual machine consolidaion for energy efficien cloud daa ceners. In: Proc. EAI In. Conf on Cloud Compuing, Oc 6. Kim SH, Kang DK, Kim WJ, Chen M, Youn CH () A science gaeway cloud wih cos adapive VM managemen for compuaional science and applicaions. o be appeared in IEEE Sys J, Chen M, Hao Y, Li Y, Lai CF, Wu D (2015) On he compuaion offloading Ad Hoc Cloudle: archiecure and service models. IEEE Commun Mag 53(6): A-Eldin A, Tordsson J, Elmroh E, Kihl M (2013) Workload classficaion for efficien auo-scaling of cloud resources. Umea Universiy, Sweden 9. Opensack, hp:// 10. Chen M, Zhana Y, Hu L, Taleb T, Shena Z (2015) Cloud-based wireless nework: virualized, reconfigurable, smar wireless nework o enable 5G echnologies. ACM/Springer Mob New Appl 20(6): Chen M, Wen Y, Jin H, Leuna V (2013) Enaling echnologies for fuure daa cener neworking: a primer. IEEE New 27(4): Xu F, Liu F, Liu L, Jin H, Li B, Li B (2014) iaware: making live migraion of virual machines inerference-aware in he cloud. IEEE Trans Compu 63(12): Gupa D, Cherkasove L, Gardner R, Vahdaa A (2006) Enforcing performance isolaion across virual machines in Xen. In: Proc. ACM/IFIP/USENIX 2006 In. Conf. Middleware, Nov 14. Nisar A, Liao WK, Choudhary A (2008) Scaling Parallel I/O Peformance hrough I/O delegae and caching sysem. In: Proc. ACM/IEEE conf on Supercompuing, Nov 15. Chen M, Zhang Y, Li Y, Mao S, Leung VCM (2015) EMC: emoion-aware mobile cloud compuing in 5G. IEEE New 29(2): Deb K, Praap A, Agarwal S, Meyarivan T (2002) A fas and eliis muliobjecive geneic algorihm:nsga-ii. IEEE Trans Evol Compu 6(2): YOCTO-WATT, hp:// 18. G-Technology, hp:// 19. PowerWake, hp://manpages.ubunu.com/manpages/uopic/man1/ powerwake.1.hml 20. Monage, hp://monage.ipac.calech.edu/ 21. Kim WJ, Kang DK, Kim SH, Youn CH (2015) Cos adapive VM managemen for scienific workflow applicaion in mobile cloud. J Mob New Appl, Springer 20(3):

Simple Network Management Based on PHP and SNMP

Simple Network Management Based on PHP and SNMP Simple Nework Managemen Based on PHP and SNMP Krasimir Trichkov, Elisavea Trichkova bsrac: This paper aims o presen simple mehod for nework managemen based on SNMP - managemen of Cisco rouer. The paper

More information

Implementing Ray Casting in Tetrahedral Meshes with Programmable Graphics Hardware (Technical Report)

Implementing Ray Casting in Tetrahedral Meshes with Programmable Graphics Hardware (Technical Report) Implemening Ray Casing in Terahedral Meshes wih Programmable Graphics Hardware (Technical Repor) Marin Kraus, Thomas Erl March 28, 2002 1 Inroducion Alhough cell-projecion, e.g., [3, 2], and resampling,

More information

MIC2569. Features. General Description. Applications. Typical Application. CableCARD Power Switch

MIC2569. Features. General Description. Applications. Typical Application. CableCARD Power Switch CableCARD Power Swich General Descripion is designed o supply power o OpenCable sysems and CableCARD hoss. These CableCARDs are also known as Poin of Disribuion (POD) cards. suppors boh Single and Muliple

More information

Scheduling. Scheduling. EDA421/DIT171 - Parallel and Distributed Real-Time Systems, Chalmers/GU, 2011/2012 Lecture #4 Updated March 16, 2012

Scheduling. Scheduling. EDA421/DIT171 - Parallel and Distributed Real-Time Systems, Chalmers/GU, 2011/2012 Lecture #4 Updated March 16, 2012 EDA421/DIT171 - Parallel and Disribued Real-Time Sysems, Chalmers/GU, 2011/2012 Lecure #4 Updaed March 16, 2012 Aemps o mee applicaion consrains should be done in a proacive way hrough scheduling. Schedule

More information

Network management and QoS provisioning - QoS in Frame Relay. . packet switching with virtual circuit service (virtual circuits are bidirectional);

Network management and QoS provisioning - QoS in Frame Relay. . packet switching with virtual circuit service (virtual circuits are bidirectional); QoS in Frame Relay Frame relay characerisics are:. packe swiching wih virual circui service (virual circuis are bidirecional);. labels are called DLCI (Daa Link Connecion Idenifier);. for connecion is

More information

Learning in Games via Opponent Strategy Estimation and Policy Search

Learning in Games via Opponent Strategy Estimation and Policy Search Learning in Games via Opponen Sraegy Esimaion and Policy Search Yavar Naddaf Deparmen of Compuer Science Universiy of Briish Columbia Vancouver, BC yavar@naddaf.name Nando de Freias (Supervisor) Deparmen

More information

User Adjustable Process Scheduling Mechanism for a Multiprocessor Embedded System

User Adjustable Process Scheduling Mechanism for a Multiprocessor Embedded System Proceedings of he 6h WSEAS Inernaional Conference on Applied Compuer Science, Tenerife, Canary Islands, Spain, December 16-18, 2006 346 User Adjusable Process Scheduling Mechanism for a Muliprocessor Embedded

More information

The Impact of Product Development on the Lifecycle of Defects

The Impact of Product Development on the Lifecycle of Defects The Impac of Produc Developmen on he Lifecycle of Rudolf Ramler Sofware Compeence Cener Hagenberg Sofware Park 21 A-4232 Hagenberg, Ausria +43 7236 3343 872 rudolf.ramler@scch.a ABSTRACT This paper invesigaes

More information

A time-space consistency solution for hardware-in-the-loop simulation system

A time-space consistency solution for hardware-in-the-loop simulation system Inernaional Conference on Advanced Elecronic Science and Technology (AEST 206) A ime-space consisency soluion for hardware-in-he-loop simulaion sysem Zexin Jiang a Elecric Power Research Insiue of Guangdong

More information

Selective Offloading in Mobile Edge Computing for the Green Internet of Things

Selective Offloading in Mobile Edge Computing for the Green Internet of Things EDGE COMPUTING FOR THE INTERNET OF THINGS Selecive Offloading in Mobile Edge Compuing for he Green Inerne of Things Xinchen Lyu, Hui Tian, Li Jiang, Alexey Vinel, Sabia Maharjan, Sein Gjessing, and Yan

More information

Optimal Crane Scheduling

Optimal Crane Scheduling Opimal Crane Scheduling Samid Hoda, John Hooker Laife Genc Kaya, Ben Peerson Carnegie Mellon Universiy Iiro Harjunkoski ABB Corporae Research EWO - 13 November 2007 1/16 Problem Track-mouned cranes move

More information

STEREO PLANE MATCHING TECHNIQUE

STEREO PLANE MATCHING TECHNIQUE STEREO PLANE MATCHING TECHNIQUE Commission III KEY WORDS: Sereo Maching, Surface Modeling, Projecive Transformaion, Homography ABSTRACT: This paper presens a new ype of sereo maching algorihm called Sereo

More information

Analysis of Various Types of Bugs in the Object Oriented Java Script Language Coding

Analysis of Various Types of Bugs in the Object Oriented Java Script Language Coding Indian Journal of Science and Technology, Vol 8(21), DOI: 10.17485/ijs/2015/v8i21/69958, Sepember 2015 ISSN (Prin) : 0974-6846 ISSN (Online) : 0974-5645 Analysis of Various Types of Bugs in he Objec Oriened

More information

Michiel Helder and Marielle C.T.A Geurts. Hoofdkantoor PTT Post / Dutch Postal Services Headquarters

Michiel Helder and Marielle C.T.A Geurts. Hoofdkantoor PTT Post / Dutch Postal Services Headquarters SHORT TERM PREDICTIONS A MONITORING SYSTEM by Michiel Helder and Marielle C.T.A Geurs Hoofdkanoor PTT Pos / Duch Posal Services Headquarers Keywords macro ime series shor erm predicions ARIMA-models faciliy

More information

A Matching Algorithm for Content-Based Image Retrieval

A Matching Algorithm for Content-Based Image Retrieval A Maching Algorihm for Conen-Based Image Rerieval Sue J. Cho Deparmen of Compuer Science Seoul Naional Universiy Seoul, Korea Absrac Conen-based image rerieval sysem rerieves an image from a daabase using

More information

Sam knows that his MP3 player has 40% of its battery life left and that the battery charges by an additional 12 percentage points every 15 minutes.

Sam knows that his MP3 player has 40% of its battery life left and that the battery charges by an additional 12 percentage points every 15 minutes. 8.F Baery Charging Task Sam wans o ake his MP3 player and his video game player on a car rip. An hour before hey plan o leave, he realized ha he forgo o charge he baeries las nigh. A ha poin, he plugged

More information

source managemen, naming, proecion, and service provisions. This paper concenraes on he basic processor scheduling aspecs of resource managemen. 2 The

source managemen, naming, proecion, and service provisions. This paper concenraes on he basic processor scheduling aspecs of resource managemen. 2 The Virual Compuers A New Paradigm for Disribued Operaing Sysems Banu Ozden y Aaron J. Goldberg Avi Silberschaz z 600 Mounain Ave. AT&T Bell Laboraories Murray Hill, NJ 07974 Absrac The virual compuers (VC)

More information

Improving the Efficiency of Dynamic Service Provisioning in Transport Networks with Scheduled Services

Improving the Efficiency of Dynamic Service Provisioning in Transport Networks with Scheduled Services Improving he Efficiency of Dynamic Service Provisioning in Transpor Neworks wih Scheduled Services Ralf Hülsermann, Monika Jäger and Andreas Gladisch Technologiezenrum, T-Sysems, Goslarer Ufer 35, D-1585

More information

Performance Evaluation of Implementing Calls Prioritization with Different Queuing Disciplines in Mobile Wireless Networks

Performance Evaluation of Implementing Calls Prioritization with Different Queuing Disciplines in Mobile Wireless Networks Journal of Compuer Science 2 (5): 466-472, 2006 ISSN 1549-3636 2006 Science Publicaions Performance Evaluaion of Implemening Calls Prioriizaion wih Differen Queuing Disciplines in Mobile Wireless Neworks

More information

COSC 3213: Computer Networks I Chapter 6 Handout # 7

COSC 3213: Computer Networks I Chapter 6 Handout # 7 COSC 3213: Compuer Neworks I Chaper 6 Handou # 7 Insrucor: Dr. Marvin Mandelbaum Deparmen of Compuer Science York Universiy F05 Secion A Medium Access Conrol (MAC) Topics: 1. Muliple Access Communicaions:

More information

NEWTON S SECOND LAW OF MOTION

NEWTON S SECOND LAW OF MOTION Course and Secion Dae Names NEWTON S SECOND LAW OF MOTION The acceleraion of an objec is defined as he rae of change of elociy. If he elociy changes by an amoun in a ime, hen he aerage acceleraion during

More information

Less Pessimistic Worst-Case Delay Analysis for Packet-Switched Networks

Less Pessimistic Worst-Case Delay Analysis for Packet-Switched Networks Less Pessimisic Wors-Case Delay Analysis for Packe-Swiched Neworks Maias Wecksén Cenre for Research on Embedded Sysems P O Box 823 SE-31 18 Halmsad maias.wecksen@hh.se Magnus Jonsson Cenre for Research

More information

An Adaptive Spatial Depth Filter for 3D Rendering IP

An Adaptive Spatial Depth Filter for 3D Rendering IP JOURNAL OF SEMICONDUCTOR TECHNOLOGY AND SCIENCE, VOL.3, NO. 4, DECEMBER, 23 175 An Adapive Spaial Deph Filer for 3D Rendering IP Chang-Hyo Yu and Lee-Sup Kim Absrac In his paper, we presen a new mehod

More information

Packet Scheduling in a Low-Latency Optical Interconnect with Electronic Buffers

Packet Scheduling in a Low-Latency Optical Interconnect with Electronic Buffers Packe cheduling in a Low-Laency Opical Inerconnec wih Elecronic Buffers Lin Liu Zhenghao Zhang Yuanyuan Yang Dep Elecrical & Compuer Engineering Compuer cience Deparmen Dep Elecrical & Compuer Engineering

More information

CENG 477 Introduction to Computer Graphics. Modeling Transformations

CENG 477 Introduction to Computer Graphics. Modeling Transformations CENG 477 Inroducion o Compuer Graphics Modeling Transformaions Modeling Transformaions Model coordinaes o World coordinaes: Model coordinaes: All shapes wih heir local coordinaes and sies. world World

More information

Voltair Version 2.5 Release Notes (January, 2018)

Voltair Version 2.5 Release Notes (January, 2018) Volair Version 2.5 Release Noes (January, 2018) Inroducion 25-Seven s new Firmware Updae 2.5 for he Volair processor is par of our coninuing effors o improve Volair wih new feaures and capabiliies. For

More information

Coded Caching with Multiple File Requests

Coded Caching with Multiple File Requests Coded Caching wih Muliple File Requess Yi-Peng Wei Sennur Ulukus Deparmen of Elecrical and Compuer Engineering Universiy of Maryland College Park, MD 20742 ypwei@umd.edu ulukus@umd.edu Absrac We sudy a

More information

Utility-Based Hybrid Memory Management

Utility-Based Hybrid Memory Management Uiliy-Based Hybrid Memory Managemen Yang Li Saugaa Ghose Jongmoo Choi Jin Sun Hui Wang Onur Mulu Carnegie Mellon Universiy Dankook Universiy Beihang Universiy ETH Zürich While he memory fooprins of cloud

More information

Difficulty-aware Hybrid Search in Peer-to-Peer Networks

Difficulty-aware Hybrid Search in Peer-to-Peer Networks Difficuly-aware Hybrid Search in Peer-o-Peer Neworks Hanhua Chen, Hai Jin, Yunhao Liu, Lionel M. Ni School of Compuer Science and Technology Huazhong Univ. of Science and Technology {chenhanhua, hjin}@hus.edu.cn

More information

A Tool for Multi-Hour ATM Network Design considering Mixed Peer-to-Peer and Client-Server based Services

A Tool for Multi-Hour ATM Network Design considering Mixed Peer-to-Peer and Client-Server based Services A Tool for Muli-Hour ATM Nework Design considering Mied Peer-o-Peer and Clien-Server based Services Conac Auhor Name: Luis Cardoso Company / Organizaion: Porugal Telecom Inovação Complee Mailing Address:

More information

PART 1 REFERENCE INFORMATION CONTROL DATA 6400 SYSTEMS CENTRAL PROCESSOR MONITOR

PART 1 REFERENCE INFORMATION CONTROL DATA 6400 SYSTEMS CENTRAL PROCESSOR MONITOR . ~ PART 1 c 0 \,).,,.,, REFERENCE NFORMATON CONTROL DATA 6400 SYSTEMS CENTRAL PROCESSOR MONTOR n CONTROL DATA 6400 Compuer Sysems, sysem funcions are normally handled by he Monior locaed in a Peripheral

More information

MOBILE COMPUTING 3/18/18. Wi-Fi IEEE. CSE 40814/60814 Spring 2018

MOBILE COMPUTING 3/18/18. Wi-Fi IEEE. CSE 40814/60814 Spring 2018 MOBILE COMPUTING CSE 40814/60814 Spring 2018 Wi-Fi Wi-Fi: name is NOT an abbreviaion play on Hi-Fi (high fideliy) Wireless Local Area Nework (WLAN) echnology WLAN and Wi-Fi ofen used synonymous Typically

More information

MOBILE COMPUTING. Wi-Fi 9/20/15. CSE 40814/60814 Fall Wi-Fi:

MOBILE COMPUTING. Wi-Fi 9/20/15. CSE 40814/60814 Fall Wi-Fi: MOBILE COMPUTING CSE 40814/60814 Fall 2015 Wi-Fi Wi-Fi: name is NOT an abbreviaion play on Hi-Fi (high fideliy) Wireless Local Area Nework (WLAN) echnology WLAN and Wi-Fi ofen used synonymous Typically

More information

IntentSearch:Capturing User Intention for One-Click Internet Image Search

IntentSearch:Capturing User Intention for One-Click Internet Image Search JOURNAL OF L A T E X CLASS FILES, VOL. 6, NO. 1, JANUARY 2010 1 InenSearch:Capuring User Inenion for One-Click Inerne Image Search Xiaoou Tang, Fellow, IEEE, Ke Liu, Jingyu Cui, Suden Member, IEEE, Fang

More information

4 Error Control. 4.1 Issues with Reliable Protocols

4 Error Control. 4.1 Issues with Reliable Protocols 4 Error Conrol Jus abou all communicaion sysems aemp o ensure ha he daa ges o he oher end of he link wihou errors. Since i s impossible o build an error-free physical layer (alhough some shor links can

More information

CAMERA CALIBRATION BY REGISTRATION STEREO RECONSTRUCTION TO 3D MODEL

CAMERA CALIBRATION BY REGISTRATION STEREO RECONSTRUCTION TO 3D MODEL CAMERA CALIBRATION BY REGISTRATION STEREO RECONSTRUCTION TO 3D MODEL Klečka Jan Docoral Degree Programme (1), FEEC BUT E-mail: xkleck01@sud.feec.vubr.cz Supervised by: Horák Karel E-mail: horak@feec.vubr.cz

More information

Video Content Description Using Fuzzy Spatio-Temporal Relations

Video Content Description Using Fuzzy Spatio-Temporal Relations Proceedings of he 4s Hawaii Inernaional Conference on Sysem Sciences - 008 Video Conen Descripion Using Fuzzy Spaio-Temporal Relaions rchana M. Rajurkar *, R.C. Joshi and Sananu Chaudhary 3 Dep of Compuer

More information

Opportunistic Flooding in Low-Duty-Cycle Wireless Sensor Networks with Unreliable Links

Opportunistic Flooding in Low-Duty-Cycle Wireless Sensor Networks with Unreliable Links 1 in Low-uy-ycle Wireless Sensor Neworks wih Unreliable Links Shuo uo, Suden Member, IEEE, Liang He, Member, IEEE, Yu u, Member, IEEE, o Jiang, Suden Member, IEEE, and Tian He, Member, IEEE bsrac looding

More information

CHANGE DETECTION - CELLULAR AUTOMATA METHOD FOR URBAN GROWTH MODELING

CHANGE DETECTION - CELLULAR AUTOMATA METHOD FOR URBAN GROWTH MODELING CHANGE DETECTION - CELLULAR AUTOMATA METHOD FOR URBAN GROWTH MODELING Sharaf Alkheder, Jun Wang and Jie Shan Geomaics Engineering, School of Civil Engineering, Purdue Universiy 550 Sadium Mall Drive, Wes

More information

Adaptive Workflow Scheduling on Cloud Computing Platforms with Iterative Ordinal Optimization

Adaptive Workflow Scheduling on Cloud Computing Platforms with Iterative Ordinal Optimization Adapive Workflow Scheduling on Cloud Compuing Plaforms wih Ieraive Ordinal Opimizaion Fan Zhang, Senior Member, IEEE; Junwei Cao, Senior Member, IEEE; Kai Hwang, Fellow, IEEE; Keqin Li, Senior Member,

More information

An efficient approach to improve throughput for TCP vegas in ad hoc network

An efficient approach to improve throughput for TCP vegas in ad hoc network Inernaional Research Journal of Engineering and Technology (IRJET) e-issn: 395-0056 Volume: 0 Issue: 03 June-05 www.irje.ne p-issn: 395-007 An efficien approach o improve hroughpu for TCP vegas in ad hoc

More information

LOW-VELOCITY IMPACT LOCALIZATION OF THE COMPOSITE TUBE USING A NORMALIZED CROSS-CORRELATION METHOD

LOW-VELOCITY IMPACT LOCALIZATION OF THE COMPOSITE TUBE USING A NORMALIZED CROSS-CORRELATION METHOD 21 s Inernaional Conference on Composie Maerials Xi an, 20-25 h Augus 2017 LOW-VELOCITY IMPACT LOCALIZATION OF THE COMPOSITE TUBE USING A NORMALIZED CROSS-CORRELATION METHOD Hyunseok Kwon 1, Yurim Park

More information

An Improved Square-Root Nyquist Shaping Filter

An Improved Square-Root Nyquist Shaping Filter An Improved Square-Roo Nyquis Shaping Filer fred harris San Diego Sae Universiy fred.harris@sdsu.edu Sridhar Seshagiri San Diego Sae Universiy Seshigar.@engineering.sdsu.edu Chris Dick Xilinx Corp. chris.dick@xilinx.com

More information

Web System for the Remote Control and Execution of an IEC Application

Web System for the Remote Control and Execution of an IEC Application Web Sysem for he Remoe Conrol and Execuion of an IEC 61499 Applicaion Oana ROHAT, Dan POPESCU Faculy of Auomaion and Compuer Science, Poliehnica Universiy, Splaiul Independenței 313, Bucureși, 060042,

More information

NRMI: Natural and Efficient Middleware

NRMI: Natural and Efficient Middleware NRMI: Naural and Efficien Middleware Eli Tilevich and Yannis Smaragdakis Cener for Experimenal Research in Compuer Sysems (CERCS), College of Compuing, Georgia Tech {ilevich, yannis}@cc.gaech.edu Absrac

More information

Delayed reservation decision in optical burst switching networks with optical buffers. Title. Li, GM; Li, VOK; Li, CY; Wai, PKA

Delayed reservation decision in optical burst switching networks with optical buffers. Title. Li, GM; Li, VOK; Li, CY; Wai, PKA Tile Delayed reservaion decision in opical burs swiching neworks wih opical buffers Auhor(s) Li, GM; Li, VOK; Li, CY; Wai, PKA Ciaion The 3rd nernaional Conference on Communicaions and Neworking in China

More information

Rule-Based Multi-Query Optimization

Rule-Based Multi-Query Optimization Rule-Based Muli-Query Opimizaion Mingsheng Hong Dep. of Compuer cience Cornell Universiy mshong@cs.cornell.edu Johannes Gehrke Dep. of Compuer cience Cornell Universiy johannes@cs.cornell.edu Mirek Riedewald

More information

Restorable Dynamic Quality of Service Routing

Restorable Dynamic Quality of Service Routing QOS ROUTING Resorable Dynamic Qualiy of Service Rouing Murali Kodialam and T. V. Lakshman, Lucen Technologies ABSTRACT The focus of qualiy-of-service rouing has been on he rouing of a single pah saisfying

More information

Test - Accredited Configuration Engineer (ACE) Exam - PAN-OS 6.0 Version

Test - Accredited Configuration Engineer (ACE) Exam - PAN-OS 6.0 Version Tes - Accredied Configuraion Engineer (ACE) Exam - PAN-OS 6.0 Version ACE Exam Quesion 1 of 50. Which of he following saemens is NOT abou Palo Alo Neworks firewalls? Sysem defauls may be resored by performing

More information

4. Minimax and planning problems

4. Minimax and planning problems CS/ECE/ISyE 524 Inroducion o Opimizaion Spring 2017 18 4. Minima and planning problems ˆ Opimizing piecewise linear funcions ˆ Minima problems ˆ Eample: Chebyshev cener ˆ Muli-period planning problems

More information

A Face Detection Method Based on Skin Color Model

A Face Detection Method Based on Skin Color Model A Face Deecion Mehod Based on Skin Color Model Dazhi Zhang Boying Wu Jiebao Sun Qinglei Liao Deparmen of Mahemaics Harbin Insiue of Technology Harbin China 150000 Zhang_dz@163.com mahwby@hi.edu.cn sunjiebao@om.com

More information

Real Time Integral-Based Structural Health Monitoring

Real Time Integral-Based Structural Health Monitoring Real Time Inegral-Based Srucural Healh Monioring The nd Inernaional Conference on Sensing Technology ICST 7 J. G. Chase, I. Singh-Leve, C. E. Hann, X. Chen Deparmen of Mechanical Engineering, Universiy

More information

SEINA: A Stealthy and Effective Internal Attack in Hadoop Systems

SEINA: A Stealthy and Effective Internal Attack in Hadoop Systems SEINA: A Sealhy and Effecive Inernal Aack in Hadoop Sysems Jiayin Wang, Teng Wang, Zhengyu Yang, Ying ao, Ningfang i, and Bo Sheng Deparmen of Compuer Science, Universiy of assachuses Boson, 1 orrissey

More information

Quick Verification of Concurrent Programs by Iteratively Relaxed Scheduling

Quick Verification of Concurrent Programs by Iteratively Relaxed Scheduling Quick Verificaion of Concurren Programs by Ieraively Relaxed Scheduling Parick Mezler, Habib Saissi, Péer Bokor, Neeraj Suri Technische Univerisä Darmsad, Germany {mezler, saissi, pbokor, suri}@deeds.informaik.u-darmsad.de

More information

Chapter 8 LOCATION SERVICES

Chapter 8 LOCATION SERVICES Disribued Compuing Group Chaper 8 LOCATION SERVICES Mobile Compuing Winer 2005 / 2006 Overview Mobile IP Moivaion Daa ransfer Encapsulaion Locaion Services & Rouing Classificaion of locaion services Home

More information

IDEF3 Process Description Capture Method

IDEF3 Process Description Capture Method IDEF3 Process Descripion Capure Mehod IDEF3 is par of he IDEF family of mehods developmen funded by he US Air Force o provide modelling suppor for sysems engineering and enerprise inegraion 2 IDEF3 Mehod

More information

Low-Cost WLAN based. Dr. Christian Hoene. Computer Science Department, University of Tübingen, Germany

Low-Cost WLAN based. Dr. Christian Hoene. Computer Science Department, University of Tübingen, Germany Low-Cos WLAN based Time-of-fligh fligh Trilaeraion Precision Indoor Personnel Locaion and Tracking for Emergency Responders Third Annual Technology Workshop, Augus 5, 2008 Worceser Polyechnic Insiue, Worceser,

More information

Let s get physical - EDA Tools for Mobility

Let s get physical - EDA Tools for Mobility Le s ge physical - EDA Tools for Mobiliy Aging and Reliabiliy Communicaion Mobile and Green Mobiliy - Smar and Safe Frank Oppenheimer OFFIS Insiue for Informaion Technology OFFIS a a glance Applicaion-oriened

More information

Page 1. Key Points from Last Lecture Frame format. EEC173B/ECS152C, Winter Wireless LANs

Page 1. Key Points from Last Lecture Frame format. EEC173B/ECS152C, Winter Wireless LANs EEC173/ECS152C, Winer 2006 Key Poins from Las Lecure Wireless LANs 802.11 Frame forma 802.11 MAC managemen Synchronizaion, Handoffs, Power MAC mehods: DCF & PCF CSMA/CA wih posiive ACK Exponenial backoff

More information

EECS 487: Interactive Computer Graphics

EECS 487: Interactive Computer Graphics EECS 487: Ineracive Compuer Graphics Lecure 7: B-splines curves Raional Bézier and NURBS Cubic Splines A represenaion of cubic spline consiss of: four conrol poins (why four?) hese are compleely user specified

More information

MATH Differential Equations September 15, 2008 Project 1, Fall 2008 Due: September 24, 2008

MATH Differential Equations September 15, 2008 Project 1, Fall 2008 Due: September 24, 2008 MATH 5 - Differenial Equaions Sepember 15, 8 Projec 1, Fall 8 Due: Sepember 4, 8 Lab 1.3 - Logisics Populaion Models wih Harvesing For his projec we consider lab 1.3 of Differenial Equaions pages 146 o

More information

Y. Tsiatouhas. VLSI Systems and Computer Architecture Lab

Y. Tsiatouhas. VLSI Systems and Computer Architecture Lab CMOS INEGRAED CIRCUI DESIGN ECHNIQUES Universiy of Ioannina Clocking Schemes Dep. of Compuer Science and Engineering Y. siaouhas CMOS Inegraed Circui Design echniques Overview 1. Jier Skew hroughpu Laency

More information

Assignment 2. Due Monday Feb. 12, 10:00pm.

Assignment 2. Due Monday Feb. 12, 10:00pm. Faculy of rs and Science Universiy of Torono CSC 358 - Inroducion o Compuer Neworks, Winer 218, LEC11 ssignmen 2 Due Monday Feb. 12, 1:pm. 1 Quesion 1 (2 Poins): Go-ack n RQ In his quesion, we review how

More information

Fully Dynamic Algorithm for Top-k Densest Subgraphs

Fully Dynamic Algorithm for Top-k Densest Subgraphs Fully Dynamic Algorihm for Top-k Denses Subgraphs Muhammad Anis Uddin Nasir 1, Arisides Gionis 2, Gianmarco De Francisci Morales 3 Sarunas Girdzijauskas 4 Royal Insiue of Technology, Sweden Aalo Universiy,

More information

A Principled Approach to. MILP Modeling. Columbia University, August Carnegie Mellon University. Workshop on MIP. John Hooker.

A Principled Approach to. MILP Modeling. Columbia University, August Carnegie Mellon University. Workshop on MIP. John Hooker. Slide A Principled Approach o MILP Modeling John Hooer Carnegie Mellon Universiy Worshop on MIP Columbia Universiy, Augus 008 Proposal MILP modeling is an ar, bu i need no be unprincipled. Slide Proposal

More information

Design Alternatives for a Thin Lens Spatial Integrator Array

Design Alternatives for a Thin Lens Spatial Integrator Array Egyp. J. Solids, Vol. (7), No. (), (004) 75 Design Alernaives for a Thin Lens Spaial Inegraor Array Hala Kamal *, Daniel V azquez and Javier Alda and E. Bernabeu Opics Deparmen. Universiy Compluense of

More information

Dynamic Route Planning and Obstacle Avoidance Model for Unmanned Aerial Vehicles

Dynamic Route Planning and Obstacle Avoidance Model for Unmanned Aerial Vehicles Volume 116 No. 24 2017, 315-329 ISSN: 1311-8080 (prined version); ISSN: 1314-3395 (on-line version) url: hp://www.ijpam.eu ijpam.eu Dynamic Roue Planning and Obsacle Avoidance Model for Unmanned Aerial

More information

Hybrid Artificial Bee Colony and Tabu Search Based Power Aware Scheduling for Cloud Computing

Hybrid Artificial Bee Colony and Tabu Search Based Power Aware Scheduling for Cloud Computing IJ Inelligen Sysems and Applicaions, 2018, 7, 39-47 Published Online July 2018 in MECS (hp://wwwmecs-pressorg/) DOI: 105815/ijisa20180704 Hybrid Arificial Bee Colony and Tabu Search Based Power Aware Scheduling

More information

Quantitative macro models feature an infinite number of periods A more realistic (?) view of time

Quantitative macro models feature an infinite number of periods A more realistic (?) view of time INFINIE-HORIZON CONSUMPION-SAVINGS MODEL SEPEMBER, Inroducion BASICS Quaniaive macro models feaure an infinie number of periods A more realisic (?) view of ime Infinie number of periods A meaphor for many

More information

Gauss-Jordan Algorithm

Gauss-Jordan Algorithm Gauss-Jordan Algorihm The Gauss-Jordan algorihm is a sep by sep procedure for solving a sysem of linear equaions which may conain any number of variables and any number of equaions. The algorihm is carried

More information

Fast Real-Time LIDAR Processing on FPGAs

Fast Real-Time LIDAR Processing on FPGAs Fas Real-Time LIDAR Processing on FPGAs K. Shih, A. Balachandran, K. Nagarajan, B. Holland, C. Slaon, A. George NSF Cener for High-Performance Reconfigurable Compuing (CHREC) Deparmen of Elecrical and

More information

Handling uncertainty in semantic information retrieval process

Handling uncertainty in semantic information retrieval process Handling uncerainy in semanic informaion rerieval process Chkiwa Mounira 1, Jedidi Anis 1 and Faiez Gargouri 1 1 Mulimedia, InfoRmaion sysems and Advanced Compuing Laboraory Sfax Universiy, Tunisia m.chkiwa@gmail.com,

More information

Announcements. TCP Congestion Control. Goals of Today s Lecture. State Diagrams. TCP State Diagram

Announcements. TCP Congestion Control. Goals of Today s Lecture. State Diagrams. TCP State Diagram nnouncemens TCP Congesion Conrol Projec #3 should be ou onigh Can do individual or in a eam of 2 people Firs phase due November 16 - no slip days Exercise good (beer) ime managemen EE 122: Inro o Communicaion

More information

I. INTRODUCTION. Keywords -- Web Server, Perceived User Latency, HTTP, Local Measuring. interchangeably.

I. INTRODUCTION. Keywords -- Web Server, Perceived User Latency, HTTP, Local Measuring. interchangeably. Evaluaing Web User Perceived Laency Using Server Side Measuremens Marik Marshak 1 and Hanoch Levy School of Compuer Science Tel Aviv Universiy, Tel-Aviv, Israel mmarshak@emc.com, hanoch@pos.au.ac.il 1

More information

Research Article An Adaptive and Integrated Low-Power Framework for Multicore Mobile Computing

Research Article An Adaptive and Integrated Low-Power Framework for Multicore Mobile Computing Hindawi Mobile Informaion Sysems Volume 2017, Aricle ID 9642958, 11 pages hps://doi.org/10.1155/2017/9642958 Research Aricle An Adapive and Inegraed Low-Power Framework for Mulicore Mobile Compuing Jongmoo

More information

Dimmer time switch AlphaLux³ D / 27

Dimmer time switch AlphaLux³ D / 27 Dimmer ime swich AlphaLux³ D2 426 26 / 27! Safey noes This produc should be insalled in line wih insallaion rules, preferably by a qualified elecrician. Incorrec insallaion and use can lead o risk of elecric

More information

M(t)/M/1 Queueing System with Sinusoidal Arrival Rate

M(t)/M/1 Queueing System with Sinusoidal Arrival Rate 20 TUTA/IOE/PCU Journal of he Insiue of Engineering, 205, (): 20-27 TUTA/IOE/PCU Prined in Nepal M()/M/ Queueing Sysem wih Sinusoidal Arrival Rae A.P. Pan, R.P. Ghimire 2 Deparmen of Mahemaics, Tri-Chandra

More information

Optimizing the Processing Performance of a Smart DMA Controller for LTE Terminals

Optimizing the Processing Performance of a Smart DMA Controller for LTE Terminals Opimizing he Processing Performance of a Smar DMA Conroller for LTE Terminals David Szczesny, Sebasian Hessel, Shadi Traboulsi, Aila Bilgic Insiue for Inegraed Sysems, Ruhr-Universiä Bochum D-78 Bochum,

More information

Chapter 2 VM Placement via Resource Brokers in a Cloud Datacenter

Chapter 2 VM Placement via Resource Brokers in a Cloud Datacenter Chapter 2 VM Placement via Resource Brokers in a Cloud Datacenter 2.1 Introduction Resource management in cloud datacenters is one of the most important issues for cloud service providers because it directly

More information

Visual Indoor Localization with a Floor-Plan Map

Visual Indoor Localization with a Floor-Plan Map Visual Indoor Localizaion wih a Floor-Plan Map Hang Chu Dep. of ECE Cornell Universiy Ihaca, NY 14850 hc772@cornell.edu Absrac In his repor, a indoor localizaion mehod is presened. The mehod akes firsperson

More information

Nonparametric CUSUM Charts for Process Variability

Nonparametric CUSUM Charts for Process Variability Journal of Academia and Indusrial Research (JAIR) Volume 3, Issue June 4 53 REEARCH ARTICLE IN: 78-53 Nonparameric CUUM Chars for Process Variabiliy D.M. Zombade and V.B. Ghue * Dep. of aisics, Walchand

More information

Chapter 4 Sequential Instructions

Chapter 4 Sequential Instructions Chaper 4 Sequenial Insrucions The sequenial insrucions of FBs-PLC shown in his chaper are also lised in secion 3.. Please refer o Chaper, "PLC Ladder diagram and he Coding rules of Mnemonic insrucion",

More information

Improved TLD Algorithm for Face Tracking

Improved TLD Algorithm for Face Tracking Absrac Improved TLD Algorihm for Face Tracking Huimin Li a, Chaojing Yu b and Jing Chen c Chongqing Universiy of Poss and Telecommunicaions, Chongqing 400065, China a li.huimin666@163.com, b 15023299065@163.com,

More information

A Progressive-ILP Based Routing Algorithm for Cross-Referencing Biochips

A Progressive-ILP Based Routing Algorithm for Cross-Referencing Biochips 16.3 A Progressive-ILP Based Rouing Algorihm for Cross-Referencing Biochips Ping-Hung Yuh 1, Sachin Sapanekar 2, Chia-Lin Yang 1, Yao-Wen Chang 3 1 Deparmen of Compuer Science and Informaion Engineering,

More information

AN ACTIVE PROGRAMMABLE HARNESS FOR MEASURMENT OF COMPOSITE NETWORK STATES

AN ACTIVE PROGRAMMABLE HARNESS FOR MEASURMENT OF COMPOSITE NETWORK STATES IEEE Inernaional Conference on Neworking, ICN 1, Colmer, France, June 1, pp68-638. AN ACTIVE PROGRAMMABLE HARNESS FOR MEASURMENT OF COMPOSITE NETWORK STATES Javed I. Khan and Asrar U. Haque Inerneworking

More information

Video-Based Face Recognition Using Probabilistic Appearance Manifolds

Video-Based Face Recognition Using Probabilistic Appearance Manifolds Video-Based Face Recogniion Using Probabilisic Appearance Manifolds Kuang-Chih Lee Jeffrey Ho Ming-Hsuan Yang David Kriegman klee10@uiuc.edu jho@cs.ucsd.edu myang@honda-ri.com kriegman@cs.ucsd.edu Compuer

More information

MoBAN: A Configurable Mobility Model for Wireless Body Area Networks

MoBAN: A Configurable Mobility Model for Wireless Body Area Networks MoBAN: A Configurable Mobiliy Model for Wireless Body Area Neworks Majid Nabi 1, Marc Geilen 1, Twan Basen 1,2 1 Deparmen of Elecrical Engineering, Eindhoven Universiy of Technology, he Neherlands 2 Embedded

More information

FIELD PROGRAMMABLE GATE ARRAY (FPGA) AS A NEW APPROACH TO IMPLEMENT THE CHAOTIC GENERATORS

FIELD PROGRAMMABLE GATE ARRAY (FPGA) AS A NEW APPROACH TO IMPLEMENT THE CHAOTIC GENERATORS FIELD PROGRAMMABLE GATE ARRAY (FPGA) AS A NEW APPROACH TO IMPLEMENT THE CHAOTIC GENERATORS Mohammed A. Aseeri and M. I. Sobhy Deparmen of Elecronics, The Universiy of Ken a Canerbury Canerbury, Ken, CT2

More information

IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS, VOL. XX, NO. XX, XX XXXX 1

IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS, VOL. XX, NO. XX, XX XXXX 1 This is he auhor's version of an aricle ha has been published in his journal. Changes were made o his version by he publisher prior o publicaion. IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS, VOL. XX,

More information

Hard Constant Bandwidth Server: Comprehensive Formulation and Critical Scenarios

Hard Constant Bandwidth Server: Comprehensive Formulation and Critical Scenarios Hard Consan Bandwidh Server: Comprehensive Formulaion and Criical Scenarios Alessandro Biondi, Alessandra Melani, Marko Berogna Scuola Superiore San Anna, Pisa, Ialy Universiy of Modena and Reggio Emilia,

More information

AN ACCESS CONTROL MODEL OF VIRTUAL MACHINE SECURITY

AN ACCESS CONTROL MODEL OF VIRTUAL MACHINE SECURITY AN ACCESS CONTROL MODEL OF VIRTUAL MACHINE SECURITY QIN Zhong-yuan 1,3, CHEN Qi 2, LV You 2, QIANG Yong 2, GUO Ai-wen 2, SHEN Ri-sheng 1,3, Zhang Qunfang 4 1 Informaion Science and Engineering School,

More information

J. Vis. Commun. Image R.

J. Vis. Commun. Image R. J. Vis. Commun. Image R. 20 (2009) 9 27 Conens liss available a ScienceDirec J. Vis. Commun. Image R. journal homepage: www.elsevier.com/locae/jvci Face deecion and racking using a Boosed Adapive Paricle

More information

Combinatorial Optimization for Embedded System Design. Luca Benini

Combinatorial Optimization for Embedded System Design. Luca Benini Combinaorial Opimizaion for Embedded Sysem Design Luca Benini Work in cooperaion wih Michela Milano s group Embedded Sysems A rough definiion Any compuing sysem which is no a compuer Large variey of devices

More information

LD7832A 4/17/2013. High Power Factor LED Controller with HV Start-up. General Description. Features. Applications. Typical Application REV: 00

LD7832A 4/17/2013. High Power Factor LED Controller with HV Start-up. General Description. Features. Applications. Typical Application REV: 00 4/17/2013 High Power Facor LED Conroller wih HV Sar-up REV: 00 General Descripion The is a buck soluion wih high PFC conrol for LED lighing. I feaures HV sar-up, easy o design wih minimum cos and PCB size.

More information

Po,,ll. I Appll I APP2 I I App3 I. Illll Illlllll II Illlll Illll Illll Illll Illll Illll Illll Illll Illll Illll Illll Illlll Illl Illl Illl

Po,,ll. I Appll I APP2 I I App3 I. Illll Illlllll II Illlll Illll Illll Illll Illll Illll Illll Illll Illll Illll Illll Illlll Illl Illl Illl Illll Illlllll II Illlll Illll Illll Illll Illll Illll Illll Illll Illll Illll Illll Illlll Illl Illl Illl US 20110153728A1 (19) nied Saes (12) Paen Applicaion Publicaion (10) Pub. No.: S 2011/0153728

More information

Automatic Calculation of Coverage Profiles for Coverage-based Testing

Automatic Calculation of Coverage Profiles for Coverage-based Testing Auomaic Calculaion of Coverage Profiles for Coverage-based Tesing Raimund Kirner 1 and Waler Haas 1 Vienna Universiy of Technology, Insiue of Compuer Engineering, Vienna, Ausria, raimund@vmars.uwien.ac.a

More information

Open Access Research on an Improved Medical Image Enhancement Algorithm Based on P-M Model. Luo Aijing 1 and Yin Jin 2,* u = div( c u ) u

Open Access Research on an Improved Medical Image Enhancement Algorithm Based on P-M Model. Luo Aijing 1 and Yin Jin 2,* u = div( c u ) u Send Orders for Reprins o reprins@benhamscience.ae The Open Biomedical Engineering Journal, 5, 9, 9-3 9 Open Access Research on an Improved Medical Image Enhancemen Algorihm Based on P-M Model Luo Aijing

More information

Motor Control. 5. Control. Motor Control. Motor Control

Motor Control. 5. Control. Motor Control. Motor Control 5. Conrol In his chaper we will do: Feedback Conrol On/Off Conroller PID Conroller Moor Conrol Why use conrol a all? Correc or wrong? Supplying a cerain volage / pulsewidh will make he moor spin a a cerain

More information

EP2200 Queueing theory and teletraffic systems

EP2200 Queueing theory and teletraffic systems EP2200 Queueing heory and eleraffic sysems Vikoria Fodor Laboraory of Communicaion Neworks School of Elecrical Engineering Lecure 1 If you wan o model neworks Or a comple daa flow A queue's he key o help

More information

SREM: A Novel Multicast Routing Algorithm - Comprehensive Cost Analysis

SREM: A Novel Multicast Routing Algorithm - Comprehensive Cost Analysis REM: A Novel Mulicas Rouing Algorihm - Comprehensive Cos Analysis Yewen Cao and Khalid Al-Begain chool of Compuing, Universiy of Glamorgan, CF37 DL, Wales, U.K E-mail:{ycao,kbegain}@glam.ac.uk Absrac Ever-increasing

More information