Location Aware Son Whitepaper Tl

The Critical Importance of Subscriber-centric Location Data for SON Use Cases

Page2

AriesoCommercialinConfidence Copyright2013AriesoLtd

Version 1.0Issued 27December2012Theinformationcontainedinthisdocumentandanydocumentationreferredtohereinorattachedhereto,isofaconfidentialnatureandissuppliedforthepurposeofdiscussiononlyandfornootherpurpose.Thisinformationshouldonlybedisclosedtothoseindividualsdirectlyinvolvedwithconsiderationandevaluationofanyproposals,allofwhoshallbemadeawareofthisrequirementforconfidentiality.Alltrademarksareherebyacknowledged.

Page3

AriesoCommercialinConfidence Copyright2013AriesoLtd

AriesoSolutionsJDSUacquiredAriesoinMarch2013,addingtheworld'sleadingintelligent,locationawaresolutionsformobilenetworkoperatorstoitsCommunicationsTestportfolio.Ariesosolutionslocate,storeandanalysedatafrombillionsofmobileconnectionevents,givingoperatorsarichsourceofintelligencetohelpboostnetworkperformanceandenrichuserexperience.Thisintelligencetransformstheeffectivenessofnetworkperformanceengineering;enablescustomercentricselfoptimisingnetworks;createstrueunderstandingofcustomerexperienceandenablesmonetizationofuniqueinsights.TheprovenAriesocarriergradesolutionsareresilientandhighlyscalable.Operatingonfivecontinents,clientsincludemobileoperatorgroupssuchasAmricaMvil,AT&T,MTN,TelefnicaandVodafone,andleadingequipmentvendorsincludingAlcatelLucentandNSN.JDSU(NASDAQ:JDSU;andTSX:JDU)innovatesandcollaborateswithcustomerstobuildandoperatethehighestperformingandhighestvaluenetworksintheworld.Ourdiversetechnologyportfolioalsofightscounterfeitingandenableshighpoweredcommerciallasersforarangeofapplications.LearnmoreaboutJDSUatwww.jdsu.comandfollowusonJDSUPerspectives,Twitter,FacebookandYouTube.MoreinformationonAriesocanbefoundatwww.arieso.com.

Page4

AriesoCommercialinConfidence Copyright2013AriesoLtd

ContentsARIESOSOLUTIONS.........................................................................................................................3EXECUTIVESUMMARY.....................................................................................................................5AUTOMATICNEIGHBOURRELATIONS..............................................................................................6COVERAGEANDCAPACITYOPTIMISATION......................................................................................8ENERGYSAVINGS............................................................................................................................9MOBILITYLOADBALANCING.........................................................................................................10MOBILITYROBUSTNESSOPTIMIZATION.........................................................................................11SUMMARY....................................................................................................................................12

Page5

AriesoCommercialinConfidence Copyright2013AriesoLtd

ExecutiveSummarySelfOptimizingNetworksofferconsiderablegainsinoperationalexpenditureefficienciesasmanynetworkimprovementtasksformerlydonemanuallybyengineerscannowbecarriedoutbyautomatedmechanisms.WhilethisisaprimarygoalofSON,therearealsoexpectationsthatSONwilloffergainsinotherareas,includingcapitalexpenditureandperformance.Inmanyregards,SONisbestconsideredastheautomationofnetworkimprovementactivitieswhichwereformerlydoneinamanualfashion.Assuch,considerationofpreviouslymanualtasksprovidesvaluableinsightintotheimportantingredientsforsuccessfulSONsolutions.Asaresultoftheintrinsicallyspatialnatureofwirelessengineering,aconsiderationoflocationinformationplaysakeyroleinnearlyallmanualoptimizationtasks.Classicexamplesincludecoverageplotsinradioplanningtools,spiderplotsinneighbourlistanalysis,andsignalstrengthplotsindrivetestpostprocessingsoftware.Therearealsomajortrendsinthewirelessindustryregardingtheuseofsubscribercentriclocationdata.ThecriticaldependencyonsubscribercentriclocationinformationcontinuesinSON,especiallyinthefollowingpopularSONusecases: AutomaticNeighbourRelations CoverageandCapacityOptimization EnergySavings MobilityLoadBalancing MobilityRobustnessOptimization

Page6

AriesoCommercialinConfidence Copyright2013AriesoLtd

AutomaticNeighbourRelationsAutomaticNeighbourRelations(ANR)istheearliestandmostextensivelydeployedSONusecasetodatearoundtheworld.ANRreferstotheabilitytogenerateusefulneighbourlistsforeachsectorinthenetwork(anexampleforonesectorisshownbelowinFigure1).Subscribercentricdataiscriticallyimportantforthegenerationofneighbourlistdetailsforinterfrequency,intrafrequencyandinterradioaccesstechnology(IRAT)applications.Thishelpstoensurethatthosesectorsthataremostoftenreportedbymobileswillbehighlyprioritizedintheensuingneighbourlistfortheservingsector.

Figure1:AutomaticNeighbourRelationsexample

Interfrequencyscenariosplaceanemphasisontheneedforlocationdataduetoexpectedvariationsinpropagationdistancesasafunctionofcarrierfrequency.ItcanalsoaccountforpowervariationsduetotheuseofdifferentRFmodulesandcabling.Subscribercentriclocationdataexplicitlyanswersthequestionofhowfarindividualcarriersextendandwhatoverlaycarriersareappropriatehandovertargetsindifferentportionsofthenetworkunderstudy.Interradioaccesstechnologyscenariosplaceanadditionalemphasisontheneedforlocationdata.Thefootprintsofeachoftheradioaccesstechnologiesunderstudymustbecarefullytakenintoconsideration(inatrafficweightedmanner)inordertoensurethatthebestIRATneighboursareidentifiedandemployed.ThistrafficweightingiskeysinceiteffectivelyallowsforapopularvotebytheactualsubscribersastowhicharethemostappropriateIRATneighbours.

Page7

AriesoCommercialinConfidence Copyright2013AriesoLtd

ExecutionofANRfunctionalityintheabsenceofsubscribercentriclocationdataposesseveralrisks.Theensuingtrial&errorsearchinawideparametricspace(whichmoreoftenthannotresemblesarandomwalk)resultsinsuboptimalnetworkoperation.Whileitis,intheory,possibletofindanoptimalcollectionofparameters,itismoreoftenthecasethatlocalmaxima(whicharesuboptimalbydefinition)willprohibitprogressalongatrajectorythatleadstotheglobalmaximum.Thisisarecurringthemeintheuseofgeographicallyblindoptimizationstrategies,aswillbenotedintheremainderofthispaper.AnotherriskassociatedwithnotusingsubscribercentriclocationdatainvolvestheimpairedabilitytoassessanddebugthesolutionsfoundbytheANRprocess.WhileengineeringinvolvementisnotexplicitlyrequiredintheheartofclosedloopSONoperations,itisstillthecasethatanySONsolutionmustbesubjecttoscrutinyanddiagnosticevaluation.Theabsenceoflocationdataprohibitstheengineerfromassessingthesituationsofsubscriberswhoarerecommendingtheaddition,deletionormaintenanceofparticularneighbours(thusprovidingacustomercentricviewofthenetwork).Forexample,locationinformationisimportantforidentificationofovershootingneighbours;itisoftenbettertoincreasetiltand/orreducetransmitpowertoeliminatetheexcessiveinterferenceoftheovershooterthantokeeptheovershootingneighbouronaneighbourlist.

Page8

AriesoCommercialinConfidence Copyright2013AriesoLtd

CoverageandCapacityOptimisationTheCoverageandCapacityOptimization(CCO)usecaseadjustsnetworkparameters(etilt,powerlevels,etc.)inordertomaximallysatisfycoverageandcapacityobjectives.InallpracticalCCOscenarios,oneofthekeyinputconstraintsisthelocationofsubscribertraffic.Thistrafficconstraintiseitherexplicitlyknowntothealgorithmcomputingtherequiredchange,orimplicit,inthatwhilstnotknowninadvance,itwillinfluencetheimpactofthechangesoncemade.UseofcustomercentriclocationdataasanexplicitinputtoCCOiscriticallyimportantbecauseitallowsfordirectconsiderationoftheconsequencesofeveryplannednetworkchangebeforeitismadeinthelivenetwork.ThisexplicitconsiderationallowstheCCOprocesstomoreeasilyarriveatoptimalnetworksolutionssuchasthoseshowninFigure2below.TheseCCOtrialresultsshowthatinthecoreofthenetwork,areaswithRSCPlevelsbelow95dBm(shownasred)havealmostdisappeared.

Figure2:CoverageandCapacityOptimizationtrialresult

AsnotedintheANRusecase,theavailabilityofsubscribercentriclocationdataallowsformoreeffectivediagnosticanalysisofCCOsolutions.Theabsenceofthisinformationmakesitnecessarytoappealtootherinferiordatasourcesinordertointerprettheconditionofthenetwork.Theseinferiordatasourcesincludedrivetestdata(whichonlyaddressroadlevelconditions)andswitchstatistics(whichonlyprovidecoarse,sectorlevelspatialresolution).

Page9

AriesoCommercialinConfidence Copyright2013AriesoLtd

EnergySavingsTheexplosionindatademandoverthepastseveralyearshasresultedintheadditionofmanycellsitesinordertosatisfyincreasingcapacityobjectivesduringhoursofpeakdemand.Thisisespeciallytrueinsituationswheretheamountoflocallyavailablespectrumisparticularlylimited.Itshouldbenotedthatthelocationswherethecapacityrelatedcellswereaddedalreadyhadadequatecoverage(asservedbythepreexistingnetworkinfrastructure).Assuch,thesenewcapacityobjectivesstandinstarkcontrasttotheoldercoverageobjectivesthatdominatedtheearlierdecadesofwirelessnetworkbuildouts.However,manyoftheserecentlyaddedsitesdonotsatisfymissioncriticalcapacityobjectivesduringoffpeakhours.Giventhattheadditionofthesenewsitesdidnotservetoimprovethelocalcoverage,itisreasonabletoexpectthatthesesites(orothermacrositesnearby)canbepowereddownduringoffpeakhoursinordertoaccomplishpowersavings.Underthesecircumstances,itisnecessarytoensurethattheoverallcoverageobjectivesarestillsatisfiedandthattheresultingnetworkstillhastherequisitecapacityasrequiredbytheactualdemandsofnearbysubscribers.Subscribercentriclocationdataiscriticallyimportanttothisusecaseinordertoensurethattheoptimalselectionofsitestobepoweredoffcanbedetermined.Suchdataalsoallowspredictionsoftheconsequencesthatwillbeseenoncethechangesarecutin.TheresultsofanEnergySavingsanalysisareshownbelowinFigure3.

Figure3:EnergySavingsanalysis(before&after)

Thisparticularexampleshowsthatthereareconsiderableopportunitiestopowerdownpartsofthenetworkduringoffpeakhourswhileensuringthatcustomersenjoythesameorbettercoveragethattheyexperienceduringpeaktrafficconditions.BasedonaTier1marketanalysis,ithasbeenfoundthatthesavingsopportunitiesvaryoveranumberofoperationalscenariosfromaconservative28%toanaggressive77%.Forasinglenetworkoperatorwith10,000sites,ithasbeenconservativelyestimatedtheproposedpowerdownstrategieswouldresultinanannualsavingsof$4.3million.Equivalentanalysesperformedwithouttheuseofcustomercentricdatainvariablyresultinaprohibitivelycomplextrial&errorwalkthroughanexponentiallylargespace.Forasfewasthirtycellsites,thereareoverabillionpossibleon/offcombinations.AssessmentoftheconsequencesofanyparticularcombinationissimilarlyconstrainedintheabsenceofcustomercentriclocationdataasnotedinearlierSONusecases.

Page10

AriesoCommercialinConfidence Copyright2013AriesoLtd

MobilityLoadBalancingTheMobilityLoadBalancing(MLB)usecaseinvolvessettingbothidleandactiveparametersinordertoensurethattrafficissuitablyspreadacrossmultipleradioaccesstechnologies.AnexampleofMLBisshowninFigure4belowwheretheredarrowsdenoteIRAThandoverbetweenLTE/UMTS,LTE/GSMandUMTS/GSMlayers.Subscribercentriclocationdataprovideskeyinputintothesettingoftheseparametersinamannerthatoptimizesthespreadoftrafficacrossthelocallyrelevantlayers,subjecttothespatialvariationsinbothsubscriberdemandaswellassubscriberdevicetype(includingwhetherdevicescanaccommodatedifferentairinterfacetechnologies).OptimalMLBstrategieswillalsotakeintoaccountthelocalspatialsupportoftheairinterfacetechnologies(similartotheANRusecasediscussionnotedearlier).

Figure4:MobilityLoadBalancingexample

LoadBalancingexercisesperformedwithouttheuseofsubscribercentriclocationdatacanoftenresultinsuboptimal,poorlydifferentiated(nearlyonesizefitsall)parametersettingsintheoutputsofMLBprocesses.ThisiscloselyrelatedtothemannerinwhichtrafficloadbalancingisaccomplishedinnonSONsystems.Atbest,thediscoveryofatrulyoptimalsolutionisgreatlydelayedbytrial&errorsearchesthroughacomplexparametricspace.DiagnosticassessmentofanyMLBoutcome(includingthedefault,onesizefitsallsetting)willalsosuffersinceother,suboptimaldatasourceswillneedtobeconsidered(drivetest,switchstatistics,etc.)Diagnosticassessmentsincludetheincreasinglyimportantchallengeofunderstandingwhycertain4Gand3Gdevicesarestrandedonlowerperformanceairinterfaces.Thesestrandedscenariosareparticularlyimportantduetothestrainthatisplacedonthecustomerexperienceandthegreatlyincreasedprobabilityofchurn.

Page11

AriesoCommercialinConfidence Copyright2013AriesoLtd

MobilityRobustnessOptimizationTheMobilityRobustnessOptimization(MRO)usecaseinvolvesoptimizationofhandoverexperiencesthroughchangesofavarietyofnetworkparameters.Suboptimalhandoverconditionsinvolveoneormoreofthefollowingsymptoms:

Tooearlyhandover Toolatehandover Unnecessaryhandover Pingponghandover Handovertothewrongcell

TheexamplebelowinFigure5showsanMROsituationwherehandoversintheredovalareoccurringinanunnecessarymanner(i.e.,wherehandoversfromthefirstcellarefollowedbyaverybriefconnectiononthesecondcell,followedbyhandoverseitherbacktothefirstcellortosomeotherthirdcell).ChangestoMROnetworkparametersresultedintheeliminationofmanyofthesehandovers(asseenintheafterpictureontheright)withoutnegativelyimpactingotheraspectsofnetworkoperation.Velocitydata(whichcanliterallybederivedfromlocationdata)canalsobeofusetooptimallydeterminetimeconstantsassociatedwithMRO.Thisisofparticularinterestinthisexamplegiventhemajoreastwestroadwayrunningthroughthemiddleoftheredoval.

Figure5:MobilityRobustnessOptimizationexample(before&after)

Page12

AriesoCommercialinConfidence Copyright2013AriesoLtd

AsnotedinearlierSONusecases,theomissionofsubscribercentriclocationdataresultsinprocessesthatrequirelengthysearchesthroughcomplexparametricspacesand/orsuboptimal,onesizefitsallparametersettings.InterpretationofanyMROsetting(includingthedefault,onesizefitsallsetting)willalsosuffersinceother,suboptimaldatasourceswillneedtobeconsidered(drivetest,switchstatistics,etc.)

SummarySelfOptimizingNetworksofferconsiderablegainsinoperationalexpenditureefficiencies,capitalexpenditureefficienciesandnetworkperformance.Theintrinsicallyspatialnatureofwirelessengineeringhasinthepastmadeconsiderationoflocationinformationthekeytosuccessinmanualoptimizationactivities.AsSONbecomesincreasinglyanprominentpartofnetworkfunction,sotheuseoflocationinformationbecomesevermorefundamental.Useofsubscribercentricsourcestoobtainlocationinformationprovidesadoublebenefit:1)Itprovidesready/relevant/inexpensiveaccesstokeydata(incontrasttodrivetesting&switchstatistics)and2)Itensuresthatthefocusremainsonthecustomer.Indeed,theneedforthislocationinformationtobederivedfromtheexperiencesofactualsubscribersisfoundtobeanaturalextensionofthesubscribercentrictrendsbeingembracedthroughoutthewirelessindustry.

Page13

AriesoCommercialinConfidence Copyright2013AriesoLtd

AriesoLtdAstorHouseNewburyBusinessParkLondonRoadNewburyBerkshireRG142PZUnitedKingdomTel: +44(0)1635232470Fax: +44(0)1635232471

AriesoInc3495PiedmontRdBldg11Suite550AtlantaGA30305USATel: +16789042424Fax: +16789042429

Email: [email protected]: www.arieso.com