BachelorThesisinSoftwareEngineering

May2013
Astudyoftheadvantages&
disadvantagesofmobilecloud
computingversusnativeenvironment
EmilAlmrot
SebastianAndersson
ContactInformation:
Authors:
EmilAlmrot
Email:ealmrot@gmail.com
SebastianAndersson
Email:sebastian.mikael.andersson@gmail.com
Universityadvisors:
KariRnkk
NinaD.Fogelstrm
SchoolofComputing Internet :www.bth.se/com
BlekingeInstituteofTechnology Phone :+46455385000
SE37179Karlskrona Fax :+46455385057
Sweden
PA1403 20130528 1
Abstract
Theadventofcloudcomputinghasenabledthepossibilityofmovingcomplexcalculationsand
deviceoperationstothecloudinaneffectiveway.Withcloudcomputingbeingappliedto
mobiledevicessomeoftheprimaryconstraintsofmobilecomputing,suchasbatterylifeand
hardwarewithlesscomputationalpower,couldberesolvedbymovingcomplexoperationsand
computationstothecloud.Thisthesisaimstoidentifyadvantagesanddisadvantages
associatedwithrunningcloudbasedapplicationsonmobiledevices.Wealsopresentastudyof
thepowerconsumptionoffivecloudbasedmobileapplicationsandcomparetheresultstotheir
noncloudcounterparts.Theresultsfromtheexperimentshowthatmigratingallyour
applicationstothecloudwillnotsignificantlyreducethepowerconsumptionofyourmobile
deviceatthemoment,butthatmobilecloudcomputinghasmaturedwithinthelastyearandwill
continuedoingsowiththedevelopmentofcloudcomputing.
PA1403 20130528 2
Table of contents
Abstract...................................................................................................................... 2
Tableofcontents....................................................................................................... 3
Introduction................................................................................................................. 4
Theory........................................................................................................................ 9
Publications......................................................................................................... 9
Relatedwork.............................................................................................................. 11
ResearchMethodology.................................................................................................. 12
Literaturestudy.................................................................................................... 13
Searchresourcesused................................................................................. 13
Searchkeywordsused.................................................................................. 13
Searchstringsused...................................................................................... 13
Searchprocess.............................................................................................. 13
EmpiricalstudyTechnicalexperiment.............................................................. 14
Results....................................................................................................................... 14
Experimentprocess............................................................................................. 14
Messaging............................................................................................................ 15
NativeMessaging(noncloud)....................................................................... 15
FacebookMessenger(cloud)........................................................................ 16
Calling.................................................................................................................. 16
NativeCalling(noncloud).............................................................................. 16
Skype(cloud)................................................................................................. 17
FileManagement.................................................................................................. 17
ESFileExplorer(noncloud).......................................................................... 18
GoogleDrive(cloud)....................................................................................... 18
Music..................................................................................................................... 19
AndroidWalkmanPlayer(noncloud)............................................................. 19
Spotify(cloud)................................................................................................. 20
VideoPlayback..................................................................................................... 20
MXPlayer(noncloud).................................................................................... 21
YouTube(cloud)............................................................................................. 21
Resultanalysis/discussion......................................................................................... 22
Conclusion.................................................................................................................. 28
References................................................................................................................. 30
Appendix..................................................................................................................... 32
Abstracts/summaries........................................................................................... 32
Keywords............................................................................................................. 36
Acronyms.............................................................................................................. 36
PA1403 20130528 3
Introduction
WiththeadventofCloudcomputing(CC),thepossibilityofmovingcomplexcalculationsand
deviceoperationstothecloud,ortheinternet,isarealpossibility.Mobiledeviceshavealways
beeninferiortoPCsintermsofhardware.Butwiththeriseofnotonlycloudcomputing,but
morespecificallymobilecloudcomputing(MCC),thereisnoreasonformobiledevicestocome
insecondplaceintermsofhardwareanymore.
[18]
Thecloudcomputingmodelintroducesathreelayerstructure,asshowninfigure1,totheway
resources,servicesandapplicationsaredistributed.ThebottomlayeristheIaaS,or
InfrastructureasaServicelayer.Thislayerrepresentsthephysicalhardwarethatprovides
computationalpower,storageetc.Thislayerispracticallyinvisibletoapplicationdevelopersand
endusersandismeanttobemanagedautomaticallyorbytheIaaSserviceprovider.
ThemiddlelayeristhePlatformasaService(PaaS)layerwhichdeliversfunctionalityto
developers.ThePaaSlayerservicesareessentiallydevelopmentplatformsthatletdevelopers
focusontheirapplicationinsteadofthephysicalhardwareproblem.
ThetoplayeristheSaaS,orSoftwareasaServicelayerwhereapplicationsinterfacewithend
users.Itisonthislayerthattheapplicationsareaccessedandusedbyitsusers.Appssuch
asYoutube,Gmail,Facebooketc.belongintheSaaSlayer.Usersdonotneedtoknowhowthe
underlyinghardwarelayer(IaaS)functionsandtheydonotneedaccesstodevelopmentand
deploymentplatformsinthePaaSlayer.
PA1403 20130528 4
Figure1:ThethreelayersofCloudComputing
Thetermcloudcomputingwasfirstusedinitscurrentcontextin1997byRamnath
Chellappa
[25]
,hesaidthatcloudcomputingwasacomputingparadigmwheretheboundariesof
computingwillbedeterminedbyeconomicrationaleratherthantechnicallimitsalone.Ithas
becomesynonymouswithecofriendlycomputingbecauseofthewayresourcesareutilized
onlyasmuchastheyareneeded.Insteadofhavingadatacenterfullofserverscapableof
handlingpeakloadrunningallday,usingupenormousamountsofpower,cloudcomputing
enablesthedynamicallocationofjustenoughresourcestocarryouttasks,whileminimizing
overhead.Loadbalancingandoffloadingtechnologiesareimperativetotheeffectivenessof
CC.
[9]
Thetermmobilecloudcomputingbeganappearingsoonaftercloudcomputinggainedsome
credibility.Itisonlylogicalthatifyoucanoffloadexpensiveoperationsfromcomputers,whynot
doitonmobiledevicesthatarebecomingincreasinglysimilartoPCs?HoweverMCC
technologyisfacedwithnotonlythestandardconcernsofcloudcomputing,buttheresalsothe
inherentproblemsofconnectivity,orlackthereof.Howdoyoubasesomethingcompletelyoffof
internetaccesswhenitcannotevenbeguaranteedinmostoftheworld?
PA1403 20130528 5
Sinceowningamobiledevicethatsconnectedtotheinternethasbecomethenorminthe
wirelessworldoftoday,MCChasaverylargepotentialuserbase.Someappsalreadyexist
whichutilizethisapproach.Oneexampleistherapidlygrowingmoviestreamingservice
Netflix
[19]
.WiththeofficialappandaNetflixservicesubscriptionitspossibletostreamanentire
movieinhighdefinitiontoamobiledevicewithoutusingupanyofthedeviceslocalstorage.This
isonlyoneexampleofhowcloudtechnologyopensuppossibilitiesthatwereneverpossible
before.
Ifmobilecloudcomputingcanfollowthetrendofregularcloudcomputingtoreceivewidespread
adoption,theseparationbetweenmobileplatformsandtheirrespectiveoperatingsystems(iOS,
Androidetc.)wouldslowlyfadeaway.Allmobiledeviceswouldaccessservicesandapplications
thesameway(e.g.byusingHTML5)andtheirhardwareperformancewouldnolongermatteras
allcomputationwouldbedoneinthecloud,onavirtualdevice,andthensentbacktotheuser.
Thiscouldalsopotentiallyprolongbatterylifeandlowerdevicecosts
[10]
,asmoreexpensive
hardwarewouldnolongerbenecessary.
Theaimofthisthesisistoevaluatewhatkindofperformanceandfeaturesonecanexpectfrom
acloudbasedapplicationonamobiledevice,whateffectsitwillhaveonthedevicethatrunsitin
termsofpowerconsumptionandperformance,andtoanalyzeissuesthatcanemergewhen
runningcloudbasedapplicationsonmobiledevices.
Twomajorconstraintsformobilecloudcomputingarelimitedenergyandwirelessbandwidth.
Studies
[8]
havebeenconductedonpowerconsumptionforcloudandnoncloudbased
applicationsandthesestudiesshowthatthecloudbasedapplicationsconsumemorepower
thanthenoncloudones.TheybelievethatthisisbecauseoftheWiFiinterfaceonmobile
devicesbeingalargefactorinpowerconsumption.Butalotofprogresshasbeenmadein
wirelessperformancesincethestudiesbyNamboodiriandGhose(2012)
[8]
andourtheoryisthat
theWiFiinterfacehasimprovedtosuchanextentthatpowerconsumptionofcloudbased
applicationsisnowmuchsmaller.
Otherconcernswithmobilecloudcomputingarehardwarewithlesscomputationalcapacityand
availability.Webelievethatthecomputationalissuecanbesolvedbymovingtheheavyliftingto
dedicatedcomputersinthecloud.Thishasbeenproventoworkinadesktopenvironmentwith
thethinclientarchitecturethatisalreadybeingusedinmanybusinessesandschools.Withthe
continueddevelopmentandexpansionof4Gwealsobelievethattheavailabilityissuecanbe
resolved.
Aswithallnewtechnologies,therewillbeproblemsandhurdlestoovercome.Thepurposeof
theevaluationistoidentifywhatproblemsthetechnologywillinitiallyface.Wewillthensurvey
ourkeyforumsforalreadyproposedsolutions,orcomeupwithourowninordertoalleviate
theseproblems.Wealsoaimtoidentifyasmanyadvantagesofmobilecloudcomputingas
possible.Ultimatelywewishtoprovethattherearenocriticaldrawbackstousingcloudbased
PA1403 20130528 6
technology,andifthereiswewanttosolvethem.
Ourresearchapproachisatechnicalexperimentfollowingourliteraturestudy.Forourliterature
studywewilluseSummon@BTHtofindrelevantpapersinthearea.Mostofthepapersusedin
thisthesisweregatheredfromtheIEEEXploredatabase.Wewillstartbyfindingpublications
aboutmobilecloudcomputingingeneralandthennarrowdownoursearchresultsbybuilding
morespecializedsearchstrings.Thepapersweultimatelywillchoosetoincludeinourthesis
willbetheoneswithharddatathatgiveussomethingtocompareourfindingsto.
Ourempiricalstudyconsistsofatechnicalexperimentwherewecomparedpowerconsumption
ofnoncloudapplicationsandtheircloudbasedcounterparts.Wecomparedthecostoftheextra
networktrafficthatisrequiredbycloudbasedappstolessnetworktrafficandlocalprocessing
thatisusedbytraditionallydesignedapps.
Seeingasthewaysmartphonesareusedvariesalotwithdemographicsandagegroups,this
meansthattherearenodefinitivetopfiveusagescenarios.InastudycarriedoutbyComscore
in2012
[22]
,itwasrevealedthatintheU.Ssendingtextmessagesisthedominantactivityon
smartphones.Thesamestudyreportsthatlisteningtomusicisalsoamongthetopactivities.
Wehaveevaluatedappsinthecategoriescalling,messaging,audioplayback,videoplayback
andfilemanagement.AnexampleisthebuiltinAndroidvideoplayerversustheYouTubeapp.
Studies
[8]
havebeenconductedonpowerconsumptionforcloudandnoncloudbased
applicationsonlaptopsandsmartphones.Thesestudiescomparespowerconsumptionofcloud
andnoncloudbasedversionsofthreetypesofapplications(wordprocessing,multimediaand
gaming).Theresultsfromthesestudiesshowthatallthreeofthecloudbasedapplications
consumemorepowerthantheirnoncloudcounterpartsonsmartphoneswhileonlythecloud
basedmultimediaapplicationconsumedmorepoweronalaptop.Boththecloudandnoncloud
basedmultimediaapplicationshadsimilarpowerconsumptionwhenrenderingthevideoonthe
devicebutthecloudbasedapplicationusedupmorepowerbecauseofthelargeamountofdata
thathadtobetransferredfromthecloud.TheybelievethisisduetothefactthattheWiFi
interfaceisamuchbiggershareofthepowerconsumptioninsmartphonesthanlaptops.
[8]
ThesepreviousexperimentshavebeendoneonoutdatedversionsoftheAndroidoperating
systems
[8]
anditcanthereforebearguedthatnetworkinterfacesofhavematuredsignificantlyin
parallelwiththegrowthofcloudsolutions.Thisiswhatweaimtofindout.
Inordertogettheresultswewerelookingfor,weidentifiedtwomainareasthatwarranted
furtherinvestigation.Theyare:
Willremotecomputation(cloudcomputing)decreasepowerconsumptiononmobile
devices?
Wewantedtoinvestigatethelikelihoodofperformancegainswhenmigratingapplicationsand
PA1403 20130528 7
operationstothecloud.Ifthereisasignificantgain,bothusersanddeveloperswillbemore
willingtoadopttothisapproach.Theresultswerethenusedtoascertainwhetheramigrationto
cloudbasedapplicationsoffersanyformofreturnofinvestment.
Arethereanydrawbacksassociatedwithrunningcloudbasedapplicationsonmobile
devicescomparedtorunningnativeones?
Arethereanyotheradvantagestorunningcloudbasedapplicationsonmobiledevices
comparedtorunningnativeones?
Ourempiricalstudyconsistedoffivecomparisonsbetweentraditionallydesigned,noncloud
basedappsandcloudbasedappswhichcouldactastheirreplacements.Tomeasurethe
powerconsumptionweusedPowerTutor
[21]
,anappdevelopedbyPhDstudentsattheuniversity
ofMichiganwiththesolepurposeofmeasuringpowerconsumptionofappsandsystem
servicesandcomponentsonAndroiddevices.Whenchoosingthepowermonitorappwe
decidedtogowiththePowerTutorappforAndroidbecauseitseemstobethedominantchoice
whenmakingpowerprofilesfortheAndroidplatform,bothfordevelopersandusers.Ithasbeen
usedbeforeinsimilarexperiments
[23]
andthewebsitepromisesaccuracywithin5%oftheactual
values
[21]
.Theonlyotheralternativeformeasuringpowerconsumptionthatwecouldfindwas
theLittleEyesoftwaresuite.WepreferredPowerTutorbecausetheLittleEyesoftwareisonly
availableforfreeasatrial,whichwefeelwouldhinderourtestssoonerorlater.Weidentified
appsfromfivemajorareasthatwefeelarethemostdesiredbytheaveragesmartphoneuser.
ThesedecisionswerebasedontheGooglePlaymarketplaceratingsatthetimeofwritingthis
thesis.Theappswechoseformeasuringthechosenusagescenarioswere:
Androidbuiltincallingappvs.Skype
AndroidbuiltinSMSappvs.Facebookmessenger
AndroidbuiltinWalkmanMusicPlayervs.Spotify
MXPlayervs.YouTube
ESFileExplorervs.GoogleDrive
Theresultsfromtheexperimentshowthatappsinsomeareashaveevolvedbetterwiththe
conceptofcloudcomputingthanothers.TheYouTubeandFacebookmessengerappsproved
toreducepowerconsumptiononthemobiledevice.TheYouTubeappsavespowerbecausethe
moreexpensiveoperationssuchasdecodingvideo/audiodataisoffloadedontoremoteservers
whilethenoncloudapp,MXPlayer,mustperformtheseoperationsonthedevice.When
comparingmessagingappswediscoveredthatthedatatransmissionoperationsarenotas
costlyassendingtextsthenormalway.ThereforetheFacebookmessengeralsoconsumes
lesspowerthanthenoncloudapp.
However,whencomparingvoicecallapps,wedeterminedthatSkypeusesalotofprocessing
powerinordertoreducebackgroundnoise.Thismakesitverybatteryheavycomparedtothe
builtincallingappinAndroidandthereforeconsumesmorepower.Ourexperimentalsoshow
thatitismorecostlytostreammusicthanplayingitlocally.Thisisbecausestreamingthedata
PA1403 20130528 8
viaanetworkconnectionismoreexpensivethanthelocalencoding/decodingprocess.
TheresultfromourfilemanagementtestshowsthatthenoncloudESFileExplorerconsumes
lesspowerthanthecloudbasedGoogleDriveappbutnotbymuch.Howeveritmustbenoted
thatESFileExploreralsoperformeditsoperationsfasterwhichmustbetakeninto
consideration.
Onahighlevelwecansaythatmobilecloudcomputingisinsomecasesapreferable
alternativetotraditionalnativeexecution.WebelievethisisduetoadvancesinbothWiFi
interfacetechnology,aswellasintheCloudcomputingfield.Comparedtotheresultsfoundby
NamboodiriandGhose(2012,pp.45)
[8]
weseethatvideoplaybackisnowsignificantlycheaper
intermsofpowerconsumption(fig.12).Ourexperimentonmessagingapplicationsshows
similarresults.
Theory
Publications
Theinformationgatheredinpapers[8],[9],[10],[11]and[12]relatetoourfirstresearch
question:
Willremotecomputation(cloudcomputing)decreasepowerconsumptionon
mobiledevices?
Theprimaryconstraintsformobilecomputingarelimitedenergyandwirelessbandwidth
[11]
and
theseconstraintsalsoapplytomobilecloudcomputing.Userstendtovaluelongerbatterylifeon
mobiledeviceshigherthanmostotherfeatures.Cloudbasedapplicationsonmobiledevices
usesatechnologyreferredtoasoffloading
[9][10]
thataimstohavemost,ifnotall,calculations
anddeviceoperationsdoneonavirtualmachineinthecloud.ThiswillresultinlessCPUusage
andmemoryconsumption
[7]
butwillitalsoreducepowerconsumptiononthedevice?
Studies
[8]
havebeenconductedonpowerconsumptionforcloudandnoncloudbased
applicationsonlaptopsandsmartphones.Thestudiescomparespowerconsumptionofcloud
andnoncloudbasedversionsofthreetypesofapplications(wordprocessing,multimediaand
gaming).Theresultsfromthesestudiesshowthatallthreeofthecloudbasedapplications
consumemorepowerthantheirnoncloudcounterpartsonsmartphoneswhileonlythecloud
basedmultimediaapplicationconsumedmorepoweronalaptop.Boththecloudandnoncloud
basedmultimediaapplicationshadsimilarpowerconsumptionwhenrenderingthevideoonthe
devicebutthecloudbasedapplicationusedupmorepowerbecauseofthelargeamountofdata
thathadtobetransferredfromthecloud.TheybelievethisisduetothefactthattheWiFi
interfaceisamuchbiggershareofthepowerconsumptioninsmartphonesthanlaptops.
[8]
Butalotofprogresshasbeenmadeinwirelessperformanceandthereareseveraltechnologies
thatarebelievedtominimizetheissuesofhighenergyconsumption.Anexampleisthe
PA1403 20130528 9
improvementsthatarebeingdonetowirelessdatatransmission.Withtheenergyefficient
wirelessdatatransmissionmechanism,thetotalenergycost(includingdatatrafficproducedby
networkrelatedapplicationsandextradatatrafficcausedbytaskoffloading)ofmobilesystems
coulddecrease.
[12]
Oneofthewaystoachievethisisthe802.11powersavingmode(PSM)
whichhasbeenwellreceivedandiswidelysupportedbywirelessinterfacecards.PSMisused
tominimizethetimespentinidlelisteninginordertoreduceenergyconsumptionandcan
thereforeprolongbatterylifeonmobiledevices.
Theinformationgatheredinpapers[13],[14],[15],[16]and[17]relatetooursecondand
thirdresearchquestions:
Arethereanydrawbacksassociatedwithrunningcloudbasedapplicationson
mobiledevicescomparedtorunningnativeones?
Arethereanyotheradvantagestorunningcloudbasedapplicationsonmobile
devicescomparedtorunningnativeones?
Withoutadoubtthebiggesthurdletoovercomeinmobilecloudcomputingisthemobilitypart.
Networkaccessandspeediscurrentlynotveryreliablewhenmovingaroundalot.Thiscould
potentiallyrendertheuseofcloudtechnologyuselessinmanypartsoftheworld.Availabilityand
QoSisstillabigproblemingeneralwhentalkingaboutCCtechnology,notonlyinthemobile
area.
[13][14]
Withthecontinueddevelopmentof4Ganditsexpansionmoreandmoreareaswill
haveinternetaccesswhichisastepintherightdirectiontosolvetheavailabilityissue.
OtherconcernswithMCCareconnectedtotheusualdrawbacksofmobiledevices.Thatis
hardwarewithlesscomputationalcapacitythantheirdesktopcounterparts.Anotheroneis
limitedbatterylife.
Thehardwareproblemwouldbemostlyalleviatedbymovingcomputationsandcomplex
operationstothecloudwherededicatedcomputerswouldcarryouttheheavylifting,andthe
mobiledevicewouldsimplysendandreceivedataoveranetwork.Thiswouldessentiallymake
mobiledevicesintothinclientsthatarealreadyprevalentinmanybusinessesandschools.
Thinclientsarelesspowerfuldesktopcomputersthatoffloadmuchoftheirworktomore
powerfulcomputerssituatedsomewhereelse.Thethinclientarchitecturewasrejuvenatedin
recentyearsinconjunctionwiththeriseofcloudcomputing.
[16]
Withtheconceptalreadyproven
toworkinadesktopenvironment,itshouldbefeasibleonmobiledevicesaswell.
OneproblemthatmobiledeviceshavethatstationaryPCsdont,istheissueofbatterylifeand
batteryconsumption.Itmoreaninherentsideeffectofmobilitythanaproblemthatcanbe
solved.
[15]
Howeveraspreviouslymentionedoffloadingcomplexoperationscouldresultina
decreasedpowerconsumptionwhichwillinturnextendbatterylife.
Anotherproblemcouldbethefixedmemoryassignedtothebrowserbythemobileoperating
system.Sincecloudbasedapplicationswillusethewebbrowserinordertorun,thiscouldlimit
theirfunctionalityandperformance.Butasthemobilecloudcomputingdevelopmentcontinues
PA1403 20130528 10
andmoreapplicationswillbemovedtothecloudthemobiledeviceswillhavemorememoryto
useandwecouldseemobileoperatingsystemsfreeupmorespacetobrowsers.
Thereisalsodifferentmobileplatformsbeingdevelopedthataimstobemoresuitableforcloud
computing.OneoftheseplatformsisCloudberry.
[17]
CloudberryisaHTML5cloudphone
platformwhichhasalltheapplications,includingthestandardapplicationssuchascontacts,
phonedialer,calendaretc.writtenaswebapplications.Alltheapplicationsarerunviaaweb
interface.Thismeansthattheseplatformsandthedevicesrunningthemcanbeoptimizedin
termsofcloudperformanceandeliminatetheissueofwebbrowsermemorylimitationsthat
othermobileoperatingsystemsmayhave.
Related work
Theinformationgatheredinpapers[1],[2],[3],[4],[5],[6]and[7]relatetoourstudyof
HTML5anditsviabilityasabasisformultiplatformcloudbasedapplications.
Ina2010Wiredmagazinearticle,ChrisAndersonandMichaelWolffprovocativelyclaimedthat
the(WorldWide)Webisdead.
[1]
Howeversince2010,alothaschangedinthewaythe(World
Wide)Webisused.TheHTML5standardcomeswithavastrangeofnewfeaturesthataimto
enrichthewebexperience.Almosttothepointoferadicatingtheneedforlocallyinstalled
applications.BelowisaquicksummaryoftheprosandconsofHTML5.
Pros
[2]
Noinstallation.
Platformindependent.(NeedsonlyaHTML5compatiblewebbrowser)
Instantdeploymentandavailability(Noneedforpublishers,distributors,appstoresetc.)
Candoalotofthethingsnormaldesktopapplicationscan.
Cons
[2]
HTML5isstillarelativelynewtechnology.
Browsersmustincreasesecuritytocopewithnewfeaturesthatcanbeused
maliciously.
Notallbrowsersareevolvingquicklyenoughtoadapttonewfunctionalityandwidespread
supportandcompatibilityisaproblemwithsomefeatures.
ManyflagshipfeaturesofHTML5continuetoblurthelinebetweendesktopapplicationandweb
page.Withgeolocation,camera,audio/videoplaybackandlocalfilemanipulationtheneedfor
installedapplicationsonthedeviceseemsmoreandmoreoutdated.
[3]
DespiteitsmanyfeaturesHTML5canstillnotcoverallaspectsoftraditionaldesktop
applications.Forinstancegameswhichrequirecomplicatedlogicandgraphicswillprobably
remainaslocallyinstalledapplicationsuntilGPUrenderinghashadtimetomatureonRIA(Rich
InternetApplication)frameworks.
[4]
PA1403 20130528 11
Whileallapplicationsmaynotbereadytofullymigratetotheweb,someofthemostglaring
problemscanbealleviatedbyusingcloudcomputing.Manybignamecompaniesarealready
rollingoutcloudsolutionsthatmakeuseofHTML5toprovideuserfriendlywebapplications.
[5]
Google,forexample,allowsuserstoattachfilestoemailsusingtheircloudstorageservice
GoogleDrive.Insteadofattachingthefilealinktothecloudstoredversionisincluded.This
reducesbothemailsizeandtransfertime.Italsogivesrecipientscontinuousaccesstothe
latest,mostuptodateversionofthefile.TheGmailappispoweredbyHTML5andtakes
advantageofmanyofitsfeatures.
TherearesignificantbenefitsinusingHTML5asthebasisforanapplication,insteadof
developingandsupportingmanydifferentversions.Thecostandeffortdecreaseishighly
desirableandtheincreasedcompatibilityandportabilityisverypositiveindeed.
[6]
The
developmentofsmallscaleappsbasedonHTML5ispromisingbutthecurrentstateofthe
standarddoesnotwarrantacompletemigrationofdesktopappstoHTML5.Howeveritshould
becomemorecommonwithinthenextyearortwo.
Research Methodology
Literature study
ThemajorityofthepaperswechosetoourliteraturestudyweregatheredfromIEEEXplore.
IEEEareoneoftheleadingdevelopersincloudcomputingandarecontinuouslyhosting
conferencesandpublishingpapersinthearea.
OursearchwasconductedbyfirstbuildingacoupleofbasicsearchstringssuchasMobile
CloudComputing,andidentifyingthekeywordsthatgaveusthemostrelevantresults.Thenwe
combinedtheseandbuiltmorespecializedsearchstringssuchasMobilecloudcomputing
offload.Themorerefinedsearchstringsgaveustheresultswewerelookingfor.Weused
Summon@BTHforthemostpartwhensearchingforpublications.Summon@BTHisa
publicationssearchenginethatoffersinstantaccesstopublicationsfromnumeroussources.It
isausefultoolwhensearchingforpublishedworkbycredibleauthors.
ThepapersweremostlysmallerpartsofscientificpublicationssuchastheInternationalJournal
ofComputerApplicationswhichcallsitselfascholarlypeerreviewedresearchpublishing
journal.Thepublicationsweusedwerescientificstudiesofdifferentareasofthecomputer
sciencefield.Theyprovideduswithharddatatosupportourclaimsandhelpedidentifyareas
wherewecouldcontribute.Sincethetechnologywearestudying(Cloudcomputing)isrelatively
new,mostofthescientificreportsthatweusedhadtobecontemporaryaswell.Themajorityof
ourreferencespointtopaperspublishedwithinthelast3years.
Thepaperswechosetoincludeinourthesisaretheoneswithharddatathatgiveussomething
tocompareourfindingsto.Wewillintentionallyleaveouttheonesthatcontainsgeneral
informationonoursubject,becausethatwillbecoveredintheintroductionandrelatedwork.
PA1403 20130528 12
Whenconductingtheempiricalinvestigationweappliedthesamelimitationsasthethingstouse
inourthesis.Wewantedhardcomparabledata,whichwefoundenoughexamplesof.Wedo
notneedthemoregeneralpapersonCloudcomputingorHTML5.
Search resources used
Google(SearchEngine)
Summon@BTH(PublicationsSearchEngine)
GoogleScholar(Referencedatabase)
BTHArkivEX(BTHpaperdatabase)
IEEEXplore(Database)
Search keywords used
Cloud
MobileCloud
CloudComputing
Offload
Performance
Wifi
Efficiency
Compatibility
HTML5
Search strings used
HTML5CloudMobile
HTML5ReplaceDesktopApplications
HTML5viability
CloudbasedHTML5
HTML5ANDdesktopORfeatures
MobilecompatibilityHTML5
Mobilecloudcomputing
MobilecloudcomputingANDperformanceORoffloadORWifi
Cloudbasedapplicationsonmobiledevices
MobileCloudOffloadingORMobileCloudComputing
MobileDeviceCloud
CloudComputingOffloadingORCloudOffloading
CloudComputing
ThinClientCloud
Search process
OurmainsourceofrelatedworkswasSummon@BTH.Ourselectionprocesswasbasedon
informationinabstractsandkeywords.
1. Firstwereadtheabstract,anddecidedwhetherthedocumentwouldcontainrelevant
PA1403 20130528 13
information.
2. Iftheabstractseemedrelevant,wecheckedthekeywordssectiontofindoutifthe
documentcontainedtheinformationwewereafter.
3. Thenextstepinvolvedscanningthroughthedocumentquicklyfortheresultschapterin
ordertovalidatethattheinformationwasworthusinginourstudy.
4. Thelaststepwassimplytoreadthroughtheentiredocumentfromstarttofinish,and
thendiscussthefindingstherein.
Empirical study - Technical experiment
Ourempiricalstudywillinvolveevaluatingdifferentcloudbasedmobileappsandcomparing
themwithtraditionallydesignedapps.Wecomparedthecostofnetworktraffic(wherecomplex
calculationsaredoneinthecloud)tolessnetworktrafficandlocalprocessing.
Wehavechosentoevaluateappsthatofferthemostcommonfunctionalitythatwethinkthe
averagesmartphoneuserdesires.Theyare:
Calling
Messaging(SMS)
Musiclistening
Videoplayback
Filemanagement
WehavefoundapplicationsthatworklocallysuchasthedefaultSMSappsoramediaplayer
thatplayslocallystoredfilesandweintendtocomparethemtoviablecloudbasedreplacements
suchastheFacebookmessengerandtheofficialYouTubeapp.Thiswillbedoneforeachofthe
listedtasks.Theresultsofthisinvestigationwillthenbeusedtoproveordisproveourtheoryof
cloudbasedapplicationsusinglesspower.
Results
Thesmartphonedeviceusedinthisexperiment:
SonyXPERIAP
Modelnumber:LT22i
AndroidOSVersion:4.0.4(ICS)
WiFiinterface:802.11/b/g/n
Inordertomeasurethepowerconsumptionoftheappsweusedanapplicationcalled
PowerTutor.
[21]
TheapplicationwascreatedbyPhDstudentsandteachersattheUniversityof
MichiganforthesolepurposeofmeasuringpowerconsumptionofdifferentappsonGoogle's
AndroidPhones.Theapplicationpromisestoprovideestimationswithin5%ofactualvalues.
Experiment process
Inordertocarryoutthepowermeasurement,makesurethatthePowerTutorapplicationis
installed.Forourexperiment,wesettheapplicationtoincludenetworktrafficandscreenactivity
aspartofthetotalpowerconsumptionofoneapplication.Eachpartoftheexperimentwas
carriedasdescribedbelow.
PA1403 20130528 14
1. StartthePowerTutorapplication.
2. Makesureallthesettingsfortheapplicationarecorrectlysetaccordingtotheabove
description.
3. Startthecaptureprocess.
4. Switchtotheapplicationbeingtested.
5. Carryoutthetaskscoveredinthescenariodescription
6. SwitchbacktothePowertutorapplicationandstopthecaptureprocess.
7. Examinethesavedlogfileandextracttherelevantdata.
8. mW=milliWatt
J=Joule
t=Timeinseconds
CalculatemWvalueusingthefollowingformula:mW=(J/t)*1000
Messaging
Thisexperimentwaschosenbecausetextingisoneofthemostcommonactivitieson
smartphonesoftoday.Wechoseasimplescenariowhere10messagesweresentand
receivedinnoparticularorder.Thismeanssometimes2messagescouldarriveaftereach
other,and3couldbesentinsuccession.Thetotalmessagecountwas10messagessent,and
10messagesreceived.
Native Messaging (non-cloud)
WiththebuiltinMessagingapplication,theexperimenttookapproximately6minutesand53
secondsintotal.Duringthefulldurationoftheexperimentthemessagingappconsumed69.5%
ofthetotalbatterypower.Itusedupatotalof176.5Jduringthetestperiod,whichequals427.36
mW.Seefigure2.
Figure2:PercentageoftotalpowerconsumptionofthebuiltinSMSapp.
Facebook Messenger (cloud)
PA1403 20130528 15
ThecloudbasedFacebookMessengerapplicationgenerallyhasfastertransfertimesof
messages(thiswouldvarydependingonthenetworkspeed),whichimpactedthetotaltime
takenfortheexperiment.Thesamesuccessionofmessageswassentoverthisapplicationin5
minutesand22seconds.Duringthisperiodtheappaccountedfor74.9%ofthetotalpower
consumptiononthephone.Itconsumedatotalof121J,whichequals375.78mW.Seefigure3.
Figure3:PercentageoftotalpowerconsumptionoftheFacebookMessengerapp.
Calling
Traditionalphonecallsisanotherobviouschoice,seeingasitisthemostbasicfunctionalityina
phone.WechosetopitthenativecallingapplicationagainsttheSkypeapp,whichisthe
dominant3rdpartyappforvoicecalls.Skypehasabadreputationofdrainingbatteryquickly.We
foundthattheseaccusationswereindeedtrue.Ourexperimentinvolvesa10minutecallwith
constantaudiodatabeingtransmitted.
Native Calling (non-cloud)
Duringtheexperiment,theappaccountedfor47.8%ofthetotalbatteryuseonthephone,and
drainedapproximately72.2J,whichisequalto120.33mW.Seefigure4.
PA1403 20130528 16
Figure4:Percentageoftotalpowerconsumptionofthebuiltincallingapp.
Skype (cloud)
Thegeneralconsensusisthatthisappisveryheavyonthebattery.Ourtestvalidatedthese
complaintsasaconversationofthesamelength(10min)required393,8J,or656,33mW.
Duringthetestperiod,theSkypeappconsumed79,6%ofthetotalbatterypowerwithinthe10
minutetimespan.Seefigure5.
Figure5:PercentageoftotalpowerconsumptionoftheSkypeapp.
File Management
Weconsiderfilemanagementtobeanotherimportanttask,withmusicandphotocollections
residingonthesamesdcard,togetherwithinstalledappsanddownloadedfiles,file
managementisanecessitywhenusingasmartphone.FortheexperimentwechosetheESFile
Explorerapp,whichseemstobethemostcommonfilemanagementapplicationontheGoogle
PA1403 20130528 17
Playmarketplacewhenwritingthisthesis.Itsupportsallthemostcommonfileoperationson
localfiles.Forthecloudbasedalternative,wechoseGoogleDriveasourcandidate.Itallows
userstomanagetheircloudbaseddrivewithease,andsupportsuploadingoffilesinadditionto
allthenormalfeatures(rename,move,etc.).
Theexperimentwasdesignedaroundwhatweconsidertobethemostcommonfileoperations.
Theyarecreating,renaming,movingandremovingfiles.Wedesignedatestscenariothat
includedallthesefeaturesinaseriesofteststeps.
1. Createafile.Nameisfile1.
2. Createafolder.Nameitfolder1.
3. Movefile1tofolder1.
4. Renamefile1tofile2.
5. Removefolder1.
ES File Explorer (non-cloud)
Theexperimenttookroughly3minutes(2minutes,54seconds)withthelocalfilemanagement
application.Duringthistime,theappconsumed40.2%ofthetotalpowerconsumption.It
consumedatotalof40.6J,or233.33mW.Seefigure6.
Figure6:PercentageoftotalpowerconsumptionoftheESFileExplorerapp.
Google Drive (cloud)
WhenusingthecloudbasedGoogleDrive,fileoperationswerealittleslower.Thetotaltimewas
3minutesand37secondsforthesamescenario.Thetestcost48.2J,or222.12mW.This
meansthattheGoogleDriveappusedlesspowerthanthelocalapp,eventhoughitneeded
longertimetocompletetheactions.duringthetest,GoogleDriveused67.5%ofthetotalbattery
power.Seefigure7.
PA1403 20130528 18
Figure7:PercentageoftotalpowerconsumptionoftheGoogleDriveapp.
Music
Playingmusicisanotheressentialfeatureofsmartphonestoday.WiththereleaseofSpotify
moreandmorepeoplestreamtheirmusicinsteadofstoringitlocally.Howeverservicessuchas
iTunesstillwarrantalocalmusiclibrarywhichmakesusthinkthatthechoicebetweenlocaland
cloudstoredmusiclibrariesisinteresting.Forthelocalmusicplayerwechosethenativeapp
whichwasinstalledonthephone,calledsimplyWalkman.Forthecloudbasedcounterpartwe
chosethemostcommonmusicstreamingserviceasof2013,whichisSpotify.
Theexperimentissimpleplaybackofasong,roughly10minutesinlength(10minutes,7
seconds)onmaxvolume.
Android Walkman Player (non-cloud)
Weexperiencedsomedifficultyinmeasuringmusicplaybackasthewalkmanappitselfonly
usedup4.3J(7.08mW),buttherearealsosystemservicescalledMediaServerandAudio
Devices,whichusedup66.5J(109.56mW)and16.4J(27.02mW)respectively.Forthesakeof
comparison,wehavedecidedtousethetotalpowerconsumptionofthesethreeinstances
combined.Thetotalpowerconsumptionwas87.2J,or143.66mWforplayingbacka10minute
trackonmaxvolumeusingthenativeWalkmanapplication.Seefigure8.
PA1403 20130528 19
Figure8:PercentageoftotalpowerconsumptionofthebuiltinWalkmanplayerapp.
Spotify (cloud)
UsingthesameapproachwhencalculatingthetotalpowerconsumptionoftheSpotifyapp,
playingbackthesame10minutesongweget127.2J(209.56mW).Ifwedividetheresultsinto
theSpotifyapp(53.6J/88.30mW),MediaServer(42.1J/69.36mW)andAudioDevices(31.5J/
51.89mW)wecanseethattheSpotifyappitselfdrawssignificantlymorepowerthanwhatthe
Walkmanappdoes.Thisismostlikelyduetothefactthatstreamingdataisamoreexpensive
operationthanopeninglocallystoredfiles.Figure9.
Figure9:PercentageoftotalpowerconsumptionoftheSpotifyapp.
Video Playback
Thelastscenariothatweconsideredtobeinterestingisvideoplayback.Whetheritsrecorded
PA1403 20130528 20
camerafootageorentiremovies,videoplaybackisanoftenusedfeatureinthesmartphonesof
today.Forthisexperimentwechosetoplaybackavideoroughly10minutesinlength(10
minutes,21seconds),in720presolutionandwithmaximumvolume.Bothplayersweresetto
keepthescreenlitduringtheentireplaybackofthefile.
MX Player (non-cloud)
WechosethemostpopularapponGooglePlayatthemomentofwritingthis,whichisMX
Player.Itplaysbacklocallystoredvideofilesinmostformats.duringtheexperiment,theapp
consumed56.5%ofthetotalbatteryconsumption.TheMediaServerserviceused23.6%ofthe
totalbatterypowerduringthetest.Theappusedupatotalof301.8J(486mW)fortheduration
ofthevideo.TheMediaServerserviceused127.9J(205.96mW).Thetotalpowerconsumption
was429.7J,or691.95mW.Seefigure10.
Figure10:PercentageoftotalpowerconsumptionoftheMXPlayerapp.
YouTube (cloud)
Forourcloudbasedalternativewechosetheleadingvideostreamingservice,Youtube,andthe
officialappforusingtheirservice.Theappaccountedfor1.4%ofthetotalbatteryconsumption
duringplayback.TheMediaServerserviceaccountedforanother21.6%ofthetotalbattery
consumption.Combinedtheyused23%.
ThetotalpowerconsumptionoftheYouTubevideoplaybackwas136.5J,or219.80mW(Media
serverincluded).Theappaloneonlyused8.4J(13.53mW),whiletheMediaServerservice
used128.1J(206.28mW).Seefigure11.
PA1403 20130528 21
Figure11:PercentageoftotalpowerconsumptionoftheYouTubeapp.
Result analysis/discussion
Inanalyzingtheresultsofourempiricalstudyofpowerefficiencyincloudcomputingwefindthat
someareashaveevolvedbetterwiththeconceptofcloudcomputing.Wecanseethatplaying
videofilesischeaperwhendonethroughstreaming,ratherthanplayingbackalocallystoredfile.
Thisisbecausethemoreexpensiveoperationssuchasdecodingvideo/audiodatacanbe
offloadedontoservers.Whenplayingbackalocalfileallthisworkhastobedoneonthedevice,
whichinturnusesmorepower.Theresultsforvideoplaybackarepresentedinfigures12and
13.
Figure12:Powerconsumptionduringvideoplayback(J).
PA1403 20130528 22
Figure13:Powerconsumptionduringvideoplayback(mW).
Theseresultsareanexampleofhowfarmobilecloudtechnologyhascomeinlessthanayear.
NamboodiriandGhose(2012,pp.45)
[8]
presentedresultsthatshowedthatcloudbasedvideo
playbackhadahigherpowerconsumption.Ourexperimentshowsasignificantdifference,
becauseinourcase,cloudbasedvideoplaybackconsumedalotlesspower.
Howeverourresultsinotherpartsoftheexperimentalsoshowthatsomeareasstillarent
matureenoughtowarrantacompleterelianceoncloudtechnology.Inmusicplaybackwesee
thatitisactuallymorecostlytostreammusicthantoplaybacklocalfiles.Thisisbecausethe
encoding/decodingprocessofsimplemp3filesischeaperthanstreamingthedataviaanetwork
connection.Therereallyisntthatmuchtogainfromstreaminglowerqualitymusic,besides
obviousreasonssuchassavingstoragespace.Theresultsforaudioplaybackarepresentedin
figures14and15.
Cuietal.
[12]
(2013,pp.4)proposeamodelforreducingthecostofdatatransmissioninmobile
cloudsolutions.Thiscouldpotentiallydecreasepowerconsumption,andinthecaseofaudio
streamingmakeitamorecosteffectivealternativethanthelocalapproach.
PA1403 20130528 23
Figure14:Powerconsumptionduringaudioplayback(J).
Figure15:Powerconsumptionduringaudioplayback(mW).
Inthecaseofvoicecalls,wehavedeterminedthattheSkypeappusesalotofprocessing
powerinordertoreducebackgroundnoiseanddeliveraudiodatawithoutstutter.Itisavery
batteryheavyappcomparedtothedefaultcallingapplicationonourtestdevice.Theaverage
usermaynotseethemoreadvancedoperationsofSkypeassomethingtheyneed,asthe
defaultcallsaremorethangoodenoughintermsofquality.Theresnoincentivetosacrificethat
muchbatterypowerunlessyoucountthefactthatSkypecallsarefree.Theresultsforvoice
callsarepresentedinfigures16and17.
PA1403 20130528 24
Figure16:Powerconsumptionforonecall(J).
Figure17:Powerconsumptionforonecall(mW).
Duringourtestsonmessaging,theresultsshowedusoncemorethatthedatatransmission
operationsarenotascostlyasonemightimagine.Inthecaseofmessaging,thesecostsare
muchlowerthanthatofsendingtextsthenormalway.WefoundthatusingtheFacebook
messengerappsavedlotsofenergy.Thisbenefit,inconjunctionwiththefactthatthefacebook
messagesarealsofreeofcharge,provesthatitisahighlydesirablesubstituteforsendingshort
textmessages.OurexperimentdidnotincludesendingimagesviaMMS/Facebookmessenger.
Theresultsformessagingarepresentedinfigures18and19.
NamboodiriandGhose(2012,pp.45)
[8]
proposesthatmobilecloudcomputingwillbeamore
PA1403 20130528 25
energyefficientalternativeif,andonlyifthereductionoflocalprocessingloadcanoffsetthe
costsofdatatransmission.Inthecaseofmessagingourresultsshowthatthecostoflocal
processingdoesoffsetthecostsofdatatransmissionmakingthecloudbasedsolutiona
preferableapproach.
Figure18:Powerconsumptionduringtextmessageconversation(J).
Figure19:Powerconsumptionduringtextmessageconversation(mW).
Ourfilemanagementtestwasaveryinterestingonebecauseasfarassoftwareandservices
goes,remotefilesstorage,andGoogleDrivehasbeenaroundforalongtime,inoneformor
PA1403 20130528 26
another.Thiswouldnaturallymakeyouassumethatsinceithashadmoretimetoevolve,it
wouldgiveclearresultsastowhatwecanexpectfrommobilecloudtechnologyasawhole.
Ourexperimentwasdesignedtoincorporatethemostbasicandcommonfileoperationswe
employasusers.Thetestcaseisdescribedinmoredetailinthepreviouschapter,Results.
Wefoundthatdespitethealreadyestablishedcloudstoragetechnology,theGoogleDriveapp
usesupslightlymorepowerwhenperformingcertainfileactionscomparedtoitslocal
counterpart.Howeveritmustbenotedthatthelocalfilemanagementapplicationwasnotonly
cheaperintermsofpower,butalsoquicker.Thisleadtotheexperimentbeingshorterwhen
usingtheESFileExplorerapp.Insteadofintentionallydelayingtheexperiment,weusedthisas
aresult.
Theconclusionisthatnotonlyisitcheaper(ifonlymarginally)tomanagefileslocally,butitis
alsofaster.
Inthecaseoffilemanagement,thecostsoflocalprocessingisnotlargeenoughtooffsetthe
transmissioncosts,thustheresultsprovethatfileoperationsarebestdonelocally.Itshouldbe
notedthatthedifferencesinpowerconsumptionareonlymarginal,andthedifferenceisbarely
noticeable.Theresultsforfilemanagementarepresentedinfigures20and21.
Figure20:Powerconsumptionduringfilemanagement(J).
PA1403 20130528 27
Figure21:Powerconsumptionduringfilemanagement(mW).
Conclusion
Followingourexperiments,wehavecometotheconclusionthatmobilecloudtechnologystill
hasroomforlotsofimprovement.Inourexperiments,thecloudbasedalternativeonlywonin2
categories,videoplaybackandmessaging.Thefilemanagementexperimentissomewhat
inconclusivebutwefindthatthedifferencebetweenmanagingfileslocallyandinthecloudis
barelynoticeable.VideoplaybackusingtheofficialYouTubeappischeaperthanplayingbacka
locallystoredfile.Thisdoesnotapplytomusicplaybackhowever,astheSpotifyappismore
costlythanplayingmp3filesfromthedevicesstorage.TheSkypeappisalsoaseriouspower
leechduringcalls,comparedtothestandardbuiltincallingapplication.Asformessaging,the
cloudbasedalternativeFacebookmessenger,ischeaperbothintermsofpowerandmoney,as
individualmessagesarefreeofcharge.
Wecanseethatmobilecloudcomputinghasevolvedsignificantlyinjustunderayearbecause
ourexperimentresultsdifferfromwhathasbeensaidpreviously.
[8][10]
Intheintroductionchapterwespecifiedthreeresearchquestions.
Willremotecomputation(cloudcomputing)decreasepowerconsumptiononmobile
devices?
Thisistruewerethereductionoflocalprocessingloadoffsetsthecostsofdatatransmissionas
proposedbyNamboodiriandGhose(2012,pp.45)
[8]
.Ourexperimentresultsshowthatpower
consumptionisdecreasedinsomecaseswhichtellsusthatcloudbasedapplicationsare
evolvingandslowlyovercomingtheirdrawbacks.Thisisespeciallyevidentinourvideoplayback
experimentwherethepowercostisfarlowerusingacloudsolution.Thiswasnotthecasea
yearago.
[8]
PA1403 20130528 28
Arethereanydrawbacksassociatedwithrunningcloudbasedapplicationsonmobile
devicescomparedtorunningnativeones?
TwomajordrawbacksofrunningcloudbasedapplicationsareavailabilityandQoS.Withthe
continueddevelopmentandexpansionof4Gwebelievethattheseissuescanberesolvedby
reducinglatencyandprovidingmoreareaswithnetworkaccess.
[14][24]
Arethereanyotheradvantagestorunningcloudbasedapplicationsonmobiledevices
comparedtorunningnativeones
Withmobilecloudcomputingallcomputationsandcomplexoperationswillbemovedtothe
cloud.Thiswilleliminatetheneedforexpensivehardwareandthereforelowerdevicecosts.It
willalsofreeupstoragespacesincemostofthedatawillbestoredinthecloudandprovide
easieraccessibility.
[14]
Asofrightnow,fullymigratingfromallofyoureverydayapplicationstotheircloudbased
counterpartswillnotresultinlongerbatterylife,asthealternativesonlydiffersignificantlyin
somecases.Whatwecansayhowever,isthatmobilecloudtechnologyhasmaturedwithinthe
lastyearortwo,andwillprobablycontinuetodosountilithassurpassedthetraditional
approach.Ascloudcomputingevolvesfurther,mobilecloudcomputingwillfollow.
PA1403 20130528 29
References
1. Mikkonen,T.Taivalsaari,A.,"ReportsoftheWebsDeathAreGreatlyExaggerated"
Computer,vol.44,no.5,pp.30,36,May2011
2. Wright,A.,ReadyforaWebOS?CommunicationsoftheACMFindingtheFunin
ComputerScienceEducation,vol.52,no.12,pp.16,17,Dec.2009
3. Hoy,Matthew.,HTML5Anewstandardfortheweb,MedicalReferenceServices
Quarterly,vol.30,no.1,pp.50,55,Jan.2011
4. Hoetzlein,R.C.,"GraphicsPerformanceinRichInternetApplications"Computer
GraphicsandApplications,vol.32,no.5,pp.98,104,Sept.Oct.2012
5. Kumar,VikasKumarGarg,Kavindra.,MigrationofServicestotheCloudEnvironment:
ChallengesandBestPractices,InternationalJournalofComputerApplications,vol.55,
no.1,pp.1,6,Oct.2012
6. Hasan,YousufZaidi,MustafaHaider,NajmiHasan,W.U.Amin,I.,SmartPhones
ApplicationdevelopmentusingHTML5andrelatedtechnologies:Atradeoffbetweencost
andquality,InternationalJournalofComputerScience,vol.9,no.3,pp.455,May2012
7. Regmi,S.S.Adhikari,S.M.S.,NetworkPerformanceofHTML5WebApplicationin
Smartphone,2011,M.Sc.,BlekingeInstituteofTechnology
8. Namboodiri,V.Ghose,T.,"Tocloudornottocloud:Amobiledeviceperspectiveon
energyconsumptionofapplications",2012IEEEInternationalSymposiumonaWorldof
Wireless,MobileandMultimediaNetworks(WoWMoM),pp.1,9,2528June
9. Kumar,KLiu,J.Lu,Y.H.Bhargava,B.,"ASurveyofComputationOffloadingfor
MobileSystems",2012,MobileNetworksandApplications,112.
10. Kumar,K.Lu,Y.H,"CloudComputingforMobileUsers:CanOffloadingComputation
SaveEnergy?",Computer,vol.43,no.4,pp.51,56,April2010
11. Altamimi,M.Palit,R.Naik,K.Nayak,A.,"EnergyasaService(EaaS):OntheEfficacy
ofMultimediaCloudComputingtoSaveSmartphoneEnergy",2012IEEE5th
InternationalConferenceonCloudComputing(CLOUD),pp.764,771,2429June2012
12. Cui,Y.Ma,X.Wang,H.Stojmenovic,I.Liu,J.,ASurveyofEnergyEfficientWireless
TransmissionandModelinginMobileCloudComputing,MobileNetworksand
Applicationsvol.18,no.1,pp.148155,2February2013
13. Pandey,A.Pooja,CloudComputingAnonDemandServicePlatform,IJCA
ProceedingsonInternationalConferenceonAdvancesinManagementandTechnology
2013,pp.5,9,February2013
14. Garg,P.Sharma,V.,ASurveyonMobileCloudComputing,IJCASpecialIssueon
IssuesandChallengesinNetworking,IntelligenceandComputingTechnologies,vol.6,
pp.15,November2012
15. NiroshinieFernando,SengW.Loke,WennyRahayu,Mobilecloudcomputing:Asurvey,
FutureGenerationComputerSystems,vol.29,no.1,pp.84,106,January2013
16. Deboosere,L.Vankeirsbilck,B.Simoens,P.DeTurck,F.Dhoedt,B.Demeester,P.,
"CloudBasedDesktopServicesforThinClients"InternetComputing,vol.16,no.6,
pp.60,67,Nov.Dec.2012
17. Taivalsaari,A.Syst,K.,"Cloudberry:AnHTML5CloudPhonePlatformforMobile
PA1403 20130528 30
Devices",Software,IEEE,vol.29,no.4,pp.40,45,JulyAug.2012
18. Dinh,H.T.,LeeC.,NiyatoD.,WangP.,2011.Asurveyofmobilecloudcomputing:
architecture,applications,andapproaches.[online]Singapore:NanyangTechnological
University.Availableat:
<http://unix.cs.plattsburgh.edu:8008/~iaydi001/COURSES/12S/12SCSC485A/READIN
GS/mobileCloudComputingWCMC2011.pdf>[Accessed21March2013]
19. Ciancutti,J.,2010.FourReasonsWeChooseAmazonsCloudasOurComputing
Platform.[blog]14December.Availableat:
<http://techblog.netflix.com/2010/12/fourreasonswechooseamazonscloudas.html>
[Accessed21March2013]
20. Song,W.XiaolongSu,"ReviewofMobilecloudcomputing"CommunicationSoftware
andNetworks(ICCSN),2011IEEE3rdInternationalConferenceon,pp.1,4,2729May
2011
21. PowerTutorAPowerMonitorforAndroidBasedMobilePlatforms
22. comScore,2012,ComscoreReportsMay2012[online]Availableat:
<http://www.comscore.com/Insights/Press_Releases/2012/7/comScore_Reports_May_
2012_U.S._Mobile_Subscriber_Market_Share>[Accessed09May2013]
23. Zhang,L.Tiwana,B.Qian,Z.Wang,Z.Dick,R.Mao,Z.M.Yang,L.,Accurateonline
powerestimationandautomaticbatterybehaviorbasedpowermodelgenerationfor
smartphonesProceedingsoftheeighthIEEE/ACM/IFIPinternationalconferenceon
Hardware/softwarecodesignandsystemsynthesis,ACM,pp.105,114,October2010
24. Cox,P.,Mobilecloudcomputing[online]Availableat:
<http://www.ibm.com/developerworks/cloud/library/clmobilecloudcomputing/>
[Accessed21May2013]
25. GoizuetaBusinessSchoolRamnathK.Chellappa
PA1403 20130528 31
Appendix
Abstracts/summaries
[1] Reports of the Webs Death Are Greatly Exaggerated
The development of entirely new types of Web-based software systems built to
leverage the vast capabilities of the World Wide Web allows the use of dynamically
downloaded applications and services from any type of terminal, including both
desktop computers and mobile devices, implying radical changes in the ways people
develop, deploy, and use software
[2] Ready for a Web OS?
A new generation of browsers may finally herald the long-awaited convergence of the
Web and operating system.
[3] HTML5 - A new standard for the web
HTML5 is the newest revision of the HTML standard developed by the World Wide Web
Consortium (W3C). This new standard adds several exciting news features and
capabilities to HTML. This article will briefly discuss the history of HTML standards,
explore what changes are in the new HTML5 standard, and what implications it has
for information professionals. A list of HTML5 resources and examples will also be
provided.
[4] Graphics Performance in Rich Internet Applications
Rendering performance for rich Internet applications (RIAs) has recently focused on
the debate between using Flash and HTML5 for streaming video and gaming on
mobile devices. A key area not widely explored, however, is the scalability of raw
bitmap graphics performance for RIAs. Does Flash render animated sprites faster
than HTML5? How much faster is WebGL than Flash? Answers to these questions are
essential for developing large-scale data visualizations, online games, and truly
dynamic websites. A new test methodology analyzes graphics performance across
RIA frameworks and browsers, revealing specific performance outliers in existing
frameworks. The results point toward a future in which all online experiences might
be GPU accelerated.
[5] Migration of Services to the Cloud Environment: Challenges and Best Practices
Cloud computing is coming-up as a new generation business paradigm with significant
potential to increase the agility and lower the costs. Business people are looking for
better ways to migrate their existing applications to a cloud-based infrastructure to
enjoy the immense cloud potential. However, the security risks, lack of mature
technology and standards, and other concerns prevent widespread enterprise
adoption of external clouds. Companies are developing a cloud computing strategy
based on growing the cloud from inside-out. Still, many applications are not suitable
for hosting at external clouds. Hence, migration of services to the external cloud
becomes a major issue whenever there is a paradigm shift. There are many key
Challenges of Enterprise Cloud Computing like Data Governance, Manageability,
Monitoring, Reliability, Availability and Virtualization. Present work discusses the major
PA1403 20130528 32
challenges towards the migration of services in cloud environment, considering both
the public as well hybrid clouds. In order to signify the challenges, both technical and
business issues have been considered. The best practices by IT giants have also
been highlighted, considering the present day business environment.
[6] Smart Phones Application development using HTML5 and related technologies:
A tradeoff between cost and quality
Smart Phone Revolution has forced companies to develop the smartphone versions of
their applications. Smart Phone market is very fragmented and volatile and there is
no single dominant platform. Building native applications for all the platforms is too
costly and time consuming. As HTML5 is emerging as a viable option for building
cross platform applications, experts are debating their quality and cost
effectiveness. In this research paper we have evaluated the quality as per ISO 9126
and look closely where HTML5 based applications are standing against the major
quality factors of ISO 9126. In this paper HTML5 based applications were evaluated
against the quality factors like Functionality, Usability, Efficiency, Maintainability,
Reliability and Portability. We have also discussed the possible cost effectiveness of
HTML5 based applications in comparison with native application development.
[7] Network Performance of HTML5 Web Application in Smartphone
Hypertext markup language 5 (HTML5), a new standard for HTML, enriched with
additional features is expected to override all the basic underlying overhead needed
by other applications. By the advent of new extension, HTML5, the webs basic
language is transplanted from a simple page layout into rich web application
development language. Furthermore, with the release of HTML5, traditional browsing
is expected to change and modify accordingly and on the other hand the potential
users will have an alternative rather than sticking in platform and OS dependent
native applications. This thesis deals with the readiness assessment of HTML5 with
regard to different smart phones- Android and Windows. In order to visualize the
fact, we analyzed different constraints like DNS lookup time, page loading time,
memory and CPU consumption associated with two applications-Flash and HTML5
running right through the smart phones. Furthermore, the comparative analysis is
performed in different network scenarios- Wi-Fi and 3G and user experience is
estimated based on network parameters. From the experiments and observations
taken, we found that android phones provide better support for HTML5 web
applications than windows mobile devices. Also, the HTML5 applications loading time
is limited by the browser rendering time rather that the content loading time from
the network and is also dependent on hardware configuration of device used.
[8] To cloud or not to cloud: A mobile device perspective on energy consumption of
applications
The cloud computing paradigm enables the work anywhere anytime paradigm by
allowing application execution and data storage on remote servers. This is especially
useful for mobile computing and communication devices that are constrained in
terms of computation power and storage. It is however not clear how preferable
cloud-based applications would be for mobile device users. For users of such battery
life constrained devices, the most important criteria might be the energy consumed
PA1403 20130528 33
by the applications they run. The goal of this work is to characterize under what
scenarios cloud-based applications would be relatively more energy-efficient for
users of mobile devices. This work first empirically studies the energy consumption
for various types of applications and for multiple classes of devices to make this
determination. Subsequently, it presents an analytical model that helps characterize
energy consumption of mobile devices under both the cloud and non-cloud
application scenarios. Finally, an algorithm GreenSpot is presented that considers
application features and energy-performance tradeoffs to determine whether cloud
or local execution will be more preferable.
[9] A Survey of Computation Offloading for Mobile Systems
Mobile systems have limited resources, such as battery life, network bandwidth,
storage capacity, and processor performance. These restrictions may be alleviated
by computation of f loading: sending heavy computation to resourceful servers and
receiving the results from these servers. Many issues related to offloading have been
investigated in the past decade. This survey paper provides an overview of the
background, techniques, systems, and research areas for offloading computation.
We also describe directions for future research.
[10] Cloud Computing for Mobile Users: Can Offloading Computation Save Energy?
The cloud heralds a new era of computing where application services are provided
through the Internet. Cloud computing can enhance the computing capability of
mobile systems, but is it the ultimate solution for extending such systems battery
lifetimes?
[11] Energy-as-a-Service (EaaS): On the Efficacy of Multimedia Cloud Computing to Save
Smartphone Energy
In spite of the dramatic growth in the number of smartphones in recent years, the
challenge of limited energy capacity of these devices has not been solved
satisfactorily. However, in the era of cloud computing, the limitation on energy
capacity can be eased off in an efcient way by ofoading heavy tasks to the cloud.
It is important for smartphone and cloud computing developers to have insights into
the energy cost of smartphone applications before implementing the ofoading
techniques. In this paper, we evaluate the energy cost of multimedia applications on
smartphones that are connected to Multimedia Cloud Computing (MCC). We have
conducted an extensive set of experiments to measure the energy costs to
investigate whether or not smartphones save energy by using MCC services. In
other words, we investigate the feasibility of MCC to provide the Energy-as-a-Service
(EaaS). Specically, we compared the energy costs for uploading and downloading a
video le to and from MCC with the energy costs of encoding the same video le on
a smartphone. The aforementioned comparison was performed by using HTTP and
FTP Internet protocols with 3G and WiFi network interfaces. All the experiments were
conducted on an Android based HTC Nexus One smartphone. Our results show
that MCC provides the smartphones with many multimedia functionalities and saves
smartphone energy from 30% to 70%.
[12] A Survey of Energy Efficient Wireless Transmission and Modeling in Mobile Cloud
Computing
PA1403 20130528 34
The emergence of mobile cloud computing (MCC) indicates that increasingly
abundant applications are available, thus deeming energy problems even more
significant. To achieve energy optimization in mobile systems, power consumption
involved with each component or application need to be estimated prior to
execution. In this paper, we present a survey on the universal energy estimation
model for mobile devices. Additionally, due to the significance of wireless network
interface card (WNIC) in the power use of mobile devices, considerable researches
have been devoted to a low-power design of the WNIC (i.e., Cellular and WiFi). These
efforts have allowed us to provide a comprehensive summary of recent work on
transmission energy savings. Finally, we conclude the survey and discuss the future
research directions.
[13] Cloud Computing - An on Demand Service Platform
Since computing required number of hardware and software resources which
continuously changed or upgraded thus requiring customers and users to
consistently pump in investments to upgrade their infrastructures. This required
researchers to arrive at a solution where users and customers would have access to
latest versions of hardware and software without requiring new investments. This led
to the very concept of cloud computing where a customer or user pays only for the
service rather than for the product. This paper is a base paper that gives a brief
introduction to the cloud computing, from what technologies cloud computing is
derived, its components, its merits and demerits, also main challenges to the cloud
computing. This paper also helps to categorize the various services provided by any
cloud service provider in some categories depending upon the type of service.
[14] A Survey on Mobile Cloud Computing
Today, during global economic downturn, exponential growth of businesses and
e-commerce deeply require better and innovative technique to reduce cost with
increasing value. Now a days due to the requirement of anytime anything from
anywhere, the volume of data accessibility increases, which gave birth to the cloud
computing. More or less Cloud computing describes highly scalable computing
resources supplied as an outer service through internet on pay-as-usability basis.
Main call for Cloud computing is that users only utilize what they required and only
pay for what they really use. Cloud computing is used to provide services to a local
client through internet on on-demand basis. With increasing use of mobile devices,
the requirement of Cloud Computing in mobile devices arises. The emanation of Cloud
Computing with Mobile web gave birth to Mobile Cloud Computing (MCC). Mobile
devices do not need to have powerful configuration because all computations can be
performed outside the mobile devices. This paper gives summary of Mobile cloud
computing (MCC), which includes introduction, architecture, applications and benefits
of MCC. It also covers what are the issues and approaches for MCC and future
research directions of Mobile Cloud Computing.
[15] Mobile cloud computing: A survey
Despite increasing usage of mobile computing, exploiting its full potential is difficult
due to its inherent problems such as resource scarcity, frequent disconnections, and
mobility. Mobile cloud computing can address these problems by executing mobile
applications on resource providers external to the mobile device. In this paper, we
PA1403 20130528 35
provide an extensive survey of mobile cloud computing research, while highlighting
the specific concerns in mobile cloud computing. We present a taxonomy based on
the key issues in this area, and discuss the different approaches taken to tackle
these issues. We conclude the paper with a critical analysis of challenges that have
not yet been fully met, and highlight directions for future work.
[16] Cloud-Based Desktop Services for Thin Clients
Cloud computing and ubiquitous network availability have renewed people's interest in
the thin client concept. By executing applications in virtual desktops on cloud
servers, users can access any application from any location with any device. For this
to be a successful alternative to traditional offline applications, however, researchers
must overcome important challenges. The thin client protocol must display
audiovisual output fluidly, and the server executing the virtual desktop should have
sufficient resources and ideally be close to the user's current location to limit
network delay. From a service provider viewpoint, cost reduction is also an
important issue.
[17] Cloudberry: An HTML5 Cloud Phone Platform for Mobile Devices
This article presents Cloudberry, a HTML5 cloud phone platform for mobile devices.
Key words
Mobile,cloud,computing,advantages,power,consumption
Acronyms
CC CloudComputing
MC MobileComputing
MCC MobileCloudComputing
SOA ServiceOrientedArchitecture
SaaS SoftwareasaService
Paas PlatformasaService
IaaS InfrastructureasaService
HTML5 HyperTextMarkupLanguage5
CPU CentralProcessingUnit
PC PersonalComputer
IEEE InstituteofElectricalandElectronicsEngineers
PSM PowerSavingMode
PA1403 20130528 36