FOG COMPUTING, what and why a fog computing

FOG COMPUTING

1. FOG
COMPUTING
BY : SHIVANGI BALI

2. Whereas the cloud is "up there" in the sky somewhere, distant and remote and
deliberately abstracted. The "fog" is close to the ground, right where things are
getting done.
Just like the cloud, fog computing is expected to open new business models.
But what a fog computing is?

3. EXISTING SYSTEM
CLOUDCOMPUTING
Inpastcoupleofyears,themainideawasgiventobuildmega
datacenterarchitecturewiththedemandofcentralized
computingservicescalledcentralizecloudcomputing(CC)
model.
Examplesofcloudcomputingare:
Google,Amazon,IBM,andMicrosoftAzure.
Cloudcomputinghasprovidedmanyopportunitiesfor
enterprisesbyofferingtheircustomerarangeofcomputing
services.Current“payasyougo”Cloudcomputingmodel
becomesanefficientalternativetoowningandmanaging
privatedatacentresforcustomersfacingWebApplications.

4. PROBLEMS WITH CLOUD COMPUTING
AstheInternetofThingsgrows,businessesfaceagrowingneedtoanalyzedatafromsourcesattheedgeofanetwork,whethermobilephones,
gateways,orIoTsensors.Cloudcomputinghasadisadvantage:
➢Itcan’tprocessdataquicklyenoughformodernbusinessapplications.TheIoTowesitsexplosivegrowthtotheconnectionofphysicalthings
andoperationtechnologies(OT)toanalyticsandmachinelearningapplications,whichcanhelpgleaninsightsfromdevice-generateddataand
enabledevicestomake“smart”decisionswithouthumanintervention.Currently,suchresourcesaremostlybeingprovidedbycloudservice
providers,wherethecomputationandstoragecapacityexists.
➢However,despiteitspower,thecloudmodelisnotapplicabletoenvironmentswhereoperationsaretime-criticalorinternetconnectivityispoor.
Thisisespeciallytrueinscenariossuchastelemedicineandpatientcare,wheremillisecondscanhavefatalconsequences.
➢Thesamecanbesaidaboutvehicletovehiclecommunications,wherethepreventionofcollisionsandaccidentscan’taffordthelatencycaused
bytheroundtriptothecloudserver.“Thecloudparadigmislikehavingyourbraincommandyourlimbsfrommilesaway—itwon’thelpyou
whereyouneedquickreflexes.”
➢Moreover,havingeverydeviceconnectedtothecloudandsendingrawdataovertheinternetcanhaveprivacy,securityandlegalimplications,
especiallywhendealingwithsensitivedatathatissubjecttoseparateregulationsindifferentcountries.IoTnodesareclosertotheaction,butfor
themoment,theydonothavethecomputingandstorageresourcestoperformanalyticsandmachinelearningtasks.Cloudservers,ontheother
hand,havethehorsepower,butaretoofarawaytoprocessdataandrespondintime.

5. PROPOSED SYSTEM
➢Fogcomputing
orfognetworking,alsoknownasfogging,isanarchitecturethatuses
edgedevicetocarryoutasubstantialamountofcomputation,storage,
andcommunicationlocallyandroutedovertheInternetbackbone.
➢Fogcomputingisadecentralizedcomputinginfrastructureinwhich
data,compute,storageandapplicationsarelocatedsomewhere
betweenthedatasourceandthecloud.Likeedgecomputing,fog
computingbringstheadvantagesandpowerofthecloudcloserto
wheredataiscreatedandactedupon.Manypeopleusethetermsfog
computingandedgecomputinginterchangeablybecausebothinvolve
bringingintelligenceandprocessingclosertowherethedatais
created.Thisisoftendonetoimproveefficiency,thoughitmightalso
bedoneforsecurityandcompliancereasons.
➢CISCOrecentlydeliveredthevisionoffogcomputingtoenable
applicationsonbillionsofconnecteddevices,alreadyconnectedin
theInternetofThings(IoT),torundirectlyatthenetworkedge.
Customerscandevelop,manageandrunsoftwareapplicationson
CiscoIOxframeworkofnetworkeddevices,includinghardened
routers,switchesandIPvideocameras.CiscoIOxbringstheopen
sourceLinuxandCiscoIOSnetworkoperatingsystemtogetherina
singlenetworkeddevice(initiallyinrouters).
Data Centres
End-Devices
Nodes

6. ARCHITECTURE
Proposed by CISCO

7. HOW FOG COMPUTING WORKS
Fognetworkingcomplements--doesn'treplace--cloudcomputing;foggingenablesshort-termanalyticsattheedge,
whilethecloudperformsresource-intensive,longer-termanalytics.
Fogcomputingusestheconceptof‘fognodes.’Thesefognodesarelocatedclosertothedatasourceandhavehigher
processingandstoragecapabilities.Fognodescanprocessthedatafarquickerthansendingtherequesttothecloud
forcentralizedprocessing.
Thecloudisgettingclutteredduetotheenormousnumberofdevicesconnectingtotheinternet.Sincecloud
computingisnotviableinsomecases,ithasbecomenecessarytousefogcomputingforIoTdevices.Itcanhandle
theenormousdatageneratedbythesedevices.
Whenimplemented,fog-empowereddeviceslocallyanalyzetime-criticaldatathatincludesalarmstatus,device
status,faultwarnings,andsoon.Thisminimizeslatencyandpreventsmajordamage.Fogcomputingcaneffectively
reducetheamountofbandwidthrequired,whichinturnspeedsupthecommunicationwiththecloudandvarious
sensors.

8. WHENTOUSEFOG COMPUTING?
FogComputingcanbeusedinthefollowingscenarios:
➢Itisusedwhenonlyselecteddataisrequiredtosendtothecloud.Thisselecteddataischosenforlongtermstorage
andislessfrequentlyaccessedbythehost.
➢Itisusedwheneveralargenumberofservicesneedtobeprovidedoveralargeareaatdifferentgeographical
locations.
➢Devicesthataresubjectedtorigorouscomputationsandprocessing'smustusefogcomputing.
➢Real-worldexampleswherefogcomputingisusedareinIoTdevices(e.g.Car-to-Carcommunication),Devices
withSensors,Cameras(IIoT-IndustrialInternetofThings),etc.

9. STEP-BY-
STEP FOG
COMPUTING
PROCESS:
Signals are wired from IoT
devices to an automation
controller which executes
a control system program
to automate those
devices.
A control system program
wires data through a
protocol gateway.
Data is converted into a
protocol such as HTTP so
that it can be understood
easily by internet-based
services.
A fog node collects the
data for further analysis.
It filters the data and saves
it for later use.

10. Cloud Architecture with and without Fog Computing
Cloud Architecture before the advent of fog computing
Cloud Architecture after the advent of fog computing

11. FOG COMPUTING CHARACTERISTICS
1.Geographicaldistribution:
Theservicesandapplication
objectiveofthefogiswidely
distributed.
2.Supportformobility:
Fogdevicesprovidemobility
techniqueslikedecouplehost
identitytolocationidentity.
3.Realtimeinteractions:
Fogcomputingrequiresreal
timeinteractionsforspeedy
service.
4. Heterogeneity:
Fog nodes can be deployed in a
wide variety of environments.
5.InteroperabilityFog
componentsmustbeableto
interoperateinordertogive
widerangeofserviceslike
streaming.
6.Decentralization:
The fog computing
architectureisdecentralized.
Thereisnocentralserverto
managecomputingresources
andservices.Therefore,fog
nodesareself-organizingand
collaboratetoprovideend
userswithreal-timeIoT
applications
Savestoragespace:
Fogcomputingisoneofthe
bestoptionstoavoidimproper
orunrelateddatatomoveto
thewholenetwork,thuswill
savestoragespaceand
decreasethelatency.

12. ADVANTAGES OF FOG COMPUTING
➢Itoffersbettersecurity.
➢FognodescanbeprotectedusingsameproceduresfollowedinITenvironment.
➢Itprocessesselecteddatalocallyinsteadofsendingthemtothecloudforprocessing.Henceitcansavenetworkbandwidth.Thisleadsto
loweroperationalcosts.
➢Itreduceslatencyrequirementsandhencequickdecisionscanbemade.Thishelpsinavoidingaccidents.
➢Itoffersbetterprivacytotheusersdataastheyareanalyzedlocallyinsteadofsendingthemtothecloud.MoreoverITteamcanmanage
andcontrolthedevices.
➢Itiseasytodevelopfogapplicationsusingrighttoolswhichcandrivemachinesaspercustomersneed.
➢Fognodesaremobileinnature.Hencetheycanjoinandleavethenetworkatanytime.
➢Fognodescanwithstandharshenvironmentalconditionsinplacessuchastracks,vehicles,undersea,factory
floorsetc.Moreoveritcanbeinstalledinremotelocations.
➢Fogcomputingoffersreductioninlatencyasdataareanalyzedlocally.Thisisduetolessroundtriptimeandlessamountofdata
bandwidth.

13. DISADVANTAGES OF FOG COMPUTING
➢Encryption algorithms and security policies make it more difficult for arbitrary devices to exchange data. Any
mistakes in security algorithms lead to exposure of data to the hackers.
➢Other security issues are IP address spoofing, man in the middle attacks, wireless network security etc.
➢To achieve high data consistency in the thefog computing is challenging and requires more efforts.
➢Fog computing will realize global storage concept with infinite size and speed of local storagebut data management is
a challenge.
➢Trust and authentication are major concerns.
➢Scheduling is complex as tasks can be moved between client devices, fog nodes and back endcloud servers.
➢Power consumption is high in fog nodes compare to centralized cloud architecture.

14. Advantages of FOG computing over CLOUD computing:
CLOUD COMPUTING FOG COMPUTING
Dataandapplicationsareprocessedinacloud,whichis
timeconsumingtaskforlargedata.
Ratherthanpresentingandworkingfromacentralized
cloud,fogoperatesonnetworkedge.Soitconsumes
lesstime.
Problemsofbandwidth,asaresultofsendingeverybit
dataovercloudcomputing
Lessdemandofbandwidth,aseverybitofdata’swere
aggregatedatcertainaccesspointinsteadofsending
overcloudchannels
Slowresponsetimeandscalabilityproblemasaresult
ofdependingserversthatarelocatedatremoteplaces.
Bysettingsmallserverscallededgeserversinvisibility
ofuser,itispossibleforafogcomputingplatformto
avoidresponsetimeandscalabilityissues

15. CHALLENGES OF FOG COMPUTING
➢Privacy:Privacyconcernisalwaystherewhentherearemanynetworksinvolved.Sincefogcomputingisbasedonwirelesstechnology,there
isahugeconcernregardingnetworkprivacy.Therearesomanyfognodesthateachend-userisaccessibletothemandbecauseofthismore
sensitiveinformationpassesfromend-userstothefognodes.
➢Security:Fogcomputingsecuritychallengeariseastherearemanydevicesconnectedtofognodesandatdifferentgateways.Eachdevicehas
adifferentIPaddress,andanyhackercanfakeyourIPaddresstogainaccesstoyourpersonalinformationthatisstoredinthatparticularfog
node.
➢FogServers:Therightplacementoffogserversshouldbetheresothatitcandeliveritsmaximumservice.Thecompanyshouldanalyzethe
demandandworkdonebythefognodebeforeplacingitwillhelpinreducingthemaintenancecost.
➢Energyconsumption:Energyconsumptionisveryhighinfogcomputingasthenumberoffognodespresentinthefogenvironmentarehigh
andrequireenergytowork.Companiesshouldtrytominimizetheenergyrequirementbythefognodessothattheyshouldbecomemore
energy-efficientandsavecosts.
➢DelayinComputing:DelaysduetoDataaggregation,Resourceover-usagereducestheeffectivesofservicesprovidedbythefogservers,
causingdelayincomputingdata.DataAggregationshouldtakeplacebeforedataprocessing,Resource-limitedfognodesshouldbedesigned
schedulingbyusingpriorityandmobilitymodel.

16. FOG COMPUTING TECHNOLOGY
➢5GTechnologies:Fogcomputingfocusesonservingcustomizedlocation-basedapplicationstomobileusers.TheFoglayers
canbeadaptedbyusingtheexistingaccessingnetworks,e.g.,WiFi,oremerging5Gwirelesstechnologieswithavirtualized
architecture.
➢NetworkFunctionVirtualization(NFV):IncontrastNFVwhichtargetstoenabledvirtualizednetworkfunctionsinside
networknodes,e.g.,switchesandrouters,Fogcomputingaimsatenablingvirtualizedlocation-basedapplicationsattheedge
deviceandprovidingdesirableservicestolocalizedmobileusers.
➢Software-definedNetworking(SDN):TheFogcomputing,asthelocalsurrogateofcloud,needstosynchronizefrequently
withcloudfordataupdateandsupport.Withaglobalnetworkview,thecloudcanmanagetheentirenetworkusingaSDN
approach.

17. APPLICATIONS
OF FOG COMPUTING

18. Smarthealthgadgetsanddevicescan
receivebestservicesfromfog
devicesatnetworkedge.Noneedto
sendservicerequesttocloudserver.
HEALTHCARE :

19. Fogdeviceenablescarstotalk
witheachothertoavoidaccidents.
Italsohelpsingettinginformation
aboutlowcrowdedpath.
CAR
COMMUNICATION :

20. Fogcomputingwouldbeableto
obtainsensordataonallthelevels,
andintegrateallthemutually
independentnetworkentities
within.
Smartcity:

21. Smartelectricdevicescanswitch
tootherenergysourceslikesolar
andwindsbasedondemandfor
energy,itsattainabilityandlow
cost.
SMART GRID:

22. LightsFogenablesautomatic
openingofsignalsbasedon
dynamicsituationoftrafficsignals.
Also,itcandetectambulanceand
openlanesforit.
SMARTTRAFFIC:

23. All sensors in the building can
communicate and exchange with each
other and with Fog device. Fog device
performs analysis on combined data.
Connected devices may react in
response to data analysis and distributed
decision making.
SMART BUILDING
Control:

24. CASE STUDEY
A REVIEW AND DISCUSSION

25. FOG COMPUTING IN HEALTHCARE
CHALLENGES FOR HEALTHCARE:
➢Healthcare system is most countries face enormous challenges that will increase due to aging population
and the rise of chronic diseases.
➢Growing nursing staff shortage.
➢Much time wasted in hospitals by manually measuring biometric parameters and transferring the data
between systems.

26. REQUIREMENTS OF HEALTHCARE APPLICATIONS
BANDWIDTH ofatleast20.48kbit/sand96kbit/s.BandwidthForECG,latenciesofupto2to4
secondsinreal-timemonitoringareacceptable.
Latency somein-bodysensorsrelyonenergy-harvesting,eitherbyheatorkineticenergy.some
sensorsmayrequireanoperationofthepatientwhenbatteryneedsreplacement.
Energy-Efficiency Dependingonwhatdataisusedfor,systemfailureshavedifferentconsequences,from
minorinconveniencetoseriousthreattothepatients’lives.
Dependability thesecurityrequirementsinhealthcarearehigh.SecuritySystems,evenwhenprovided
bydifferentvendors,shouldbeinteroperablewitheachother.
Interoperability system even when provided by the different vendors, should be interoperable with each
other.

27. VISION OF FOG COMPUTING IN HEALTHCARE
➢FlexibilityofComputationlocus:Thelocationcanbedynamicanddependonthecurrentcontext,environment
andapplicationrequirements.
➢Integration:Withinfogcomputingarchitecture,newsensorscanbeaddedtotheexistinginfrastructure.
•Fogcomputingcanalsoserveasacompatibilitylayertotranslatebetweenvariousstandards.
➢Patientmobility:Application-specificinfrastructurealsolimitstheareawherepatientscanbemonitored.
•Thetransitionsbetweendifferentenvironmentscanbemanagedmoregradually.
➢NewApplications:Fogcomputingwillprovidelatencyandresponsetimeimprovements,aswellasenergy
savingsforwearableandlow-costdevices,whileperformingcomplextaskssuchasfalldetection.
•InternetofHealthcarethings.

29. Healthcare Application with and without Fog Computing
WITHOUT FOG COMPUTING WITH FOG COMPUTING

30. SECURITY ISSUES
Themainsecurityissuesareauthenticationatdifferentlevelsofgatewaysaswellasincaseofsmartgridsatthe
smartmetersinstalledintheconsumer’shome.EachsmartmeterandsmartappliancehasanIPaddress.A
malicioususercaneithertamperwithitsownsmartmeter,reportfalsereadings,orspoofIPaddresses.

31. Example: MAN-IN –MIDDLE-ATTACK
In this subsection, we take man-in-the-middle attack as an example to expose the security problems in Fog
computing. In this attack, gateways serving as Fog devices may be compromised or replaced by fake ones .

32. CONCLUSION
Fogcomputingadvantagesforservicesinseveraldomains,such
asSmartGrid,wirelesssensornetworks,InternetofThings(IoT)
andsoftwaredefinednetworks(SDNs).Weexaminethestate-
of-the-artanddisclosesomegeneralissuesinFogcomputing
includingsecurity,privacy,trust,andservicemigrationamong
FogdevicesandbetweenFogandCloud.

33. FUTURE ENCHANCEMENT
Since fog as itself a newer concept, implanting this will takes a little time but once done, it would make IoT devices a lot smarter.
The extended work on fog computing & cloud computing could be on the impact of heterogeneous storage and overall
performance on the basis ofdiverse applications.
Fog computing has several advantages overcloud computing. Fog computing can boost usability and accessibility in various
computing environments. Soon,cloud computing for IoTmay fade away but fog computing will take over. IoT is seeing an
impressive growth rate and so it needs a special infrastructure base that can handle all its requirements. Fog computing is the key
to accomplish this critical work.

34. THANK YOU