CONTENTS

1.ExecutiveSummary.......................................................................................................................................3 2.Introduction..................................................................................................................................................4 2.1TheMobileMiracle....................................................................................................................................... 4 2.2WhyItMustbeNurtured............................................................................................................................... 5 3.SpectrumforMobileBroadband....................................................................................................................7 3.1.SpectrumDemand......................................................................................................................................... 7 3.2SpectrumRequirements............................................................................................................................. 6 1 3.3TakeAways................................................................................................................................................. 1 2 4.SpectrumAnimatingPrinciples..................................................................................................................... 7 2 4.1Wellconceivedspectrumstrategiesarevital............................................................................................. 7 2 4.2Marketorientedassignmentapproachesworkspectrumcapsshouldbedisfavored............................ 0 4 4.3Thereisnotimetolosespectrumallocationscantakeyearstoeffectuate........................................... 4 4 5.ClosingConsiderations.................................................................................................................................. 6 4 6.Appendices.................................................................................................................................................. 7 4 6.1Abbreviations.............................................................................................................................................. 7 4 6.2SelectedReferences.................................................................................................................................... 9 4 7.Acknowledgements...................................................................................................................................... 0 5

1.EXECUTIVESUMMARY

Mobile broadband in the Americas is in a delicate state. On the one hand, the growth in subscribershiphasbeenphenomenalamobilemiracle,astheITUcharacterizeditin2008, when mobile broadband subscriptions first exceeded fixed broadband subscriptions globally. Thefeaturesandcapabilitiesoftodaysmobilebroadbandnetworks,devicesandservicesare astounding,withthepromiseoffutureenhancementsevenmoreexciting. On the other hand, the mobile broadband stands at a potentially perilous time. Today the industry lacks sufficient incremental supply of one of its essential raw materials spectrum. Based on a review of recent forecasts, one can reasonably draw the conclusion that mobile broadbandnetworkswillhitcapacityshortagesbythemiddleofthedecadeunlessstepsare takentosecuretheadditionalspectrumneeded. Such steps need to be taken today to avoid these risks. Helpfully, the industry has a long history of driving innovation in radio access technologies, from EDGE through HSPA to LTE, allowingittoexploitspectrumassetsasintensivelyaspossible.Inparallel,theindustryalsohas invested billions in building cell sites to enhance network coverage and capacity. Such steps willcontinuetobeneeded,andthereisnoindicationofdeploymentslowing. Atthesametime,additionalspectrumformobilebroadbandisvital.Countriesmustbeginnow to plan for the future, in order to preserve the promise of the mobile miracle. Historically, spectrumallocationscantakeatleast5years,andoftenlonger,toeffectuate. 4G Americas offers the following guideposts to aid stakeholders in the region in working togethertosecureabrightmobilebroadbandtomorrow. 1. WellConsideredSpectrumAllocationPoliciesareImperative A. ConfigureLicenseswithWiderBandwidths B. GroupLikeServicesTogether C. BeMindfulofGlobalTechnologyStandards D. PursueHarmonized/ContiguousSpectrumAllocations E. ExhaustExclusiveUseOptionsBeforePursuingSharedUse F. NotAllSpectrumisFungibleAlignAllocationwithDemand 2.Marketorientedspectrumassignmentapproachesworkspectrumcapsshouldbe disfavored. 3.Thereisnotimetolosespectrumallocationscantakeyearstoeffectuate. 4GAmericasmembers,aswellasitsmembercompanies,standpreparedtoaidstakeholdersin securingthepromiseofmobilebroadband.

3|P a g e

2.INTRODUCTION

2.1THEMOBILEMIRACLE It is nothing short of a mobile miracle. So remarked the International Telecommunications Union(ITU),anagencyoftheUnitedNations,inaMay2010reportonmeasuringtheglobal Information Society. Taking note of the astounding developments in the growth of mobile broadband,theITUstatedthat: Promisingdevelopmentsarecurrentlytakingplaceinthemobile broadbandsector.TheintroductionofhighspeedmobileInternet accessinanincreasingnumberofcountrieswillfurtherboostthe number of Internet users, particularly in the developing world. Indeed,thenumberofmobilebroadbandsubscriptionshasgrown steadilyandin2008surpassedthoseforfixedbroadband.Atthe endof2009,therewereanestimated640millionmobileand490 millionfixedbroadbandsubscriptions. TheITUhighlightedthatmobilebroadbandsubscriptionsnowmateriallysurpassfixed broadbandsubscriptions,andprovidedthefollowinggraphicpunctuatingthisinflectionpoint:
4|P a g e

Prior to delving into these trends in more depth, it is instructive to pause to reflect on the significance of these trends for our society and its development. Specifically, why is it importanttonurturemobilebroadband1deployment? 2.2WHYITMUSTBENURTURED As the United Nations Conference on Trade and Development (UNCTAD) succinctly phrased matters, the benefits of broadband for social and economic development are well acknowledged.2 It references the frequently cited World Bank finding that, in low and middleincome countries, every ten percentage point increase in broadband penetration correspondstoanincreaseineconomicgrowthof1.38percentagepoints;doublethatofhigh incomecountries. Morepreciselyforpresentpurposes,UNCTADnotesthat Wireless offers a more practical broadband entry point for developing nations. Installation costs are lower than fixed broadband and for mobile broadband,countriescanleverageexistingnetworks. MostnationsinLatinAmericaandtheCaribbeanarefacedwiththeurgentneedtoreducethe digital divide by increasing broadband penetration. At best, broadband penetration hovers aroundtenpercent.Incontrast,wirelesspenetrationforbasicvoiceanddataservicesalready reaches95%intheregion,andhassurpassed100%inseveralcountries. Asaresult,mobilebroadbandislikelytheoptimalsometimestheonlyenablerofInternet connectivity for residents of the region. Encouragingly, in burgeoning Latin American economies such as Argentina, Brazil, Chile, Colombia, Mexico and Uruguay, growth in mobile broadbandhasexceededtherateofbroadbandaccessprovidedbyfixednetworksduringthe lasttwelvemonths. Mobile broadband is already contributing significantly to the GDPs of emerging countries. Globally,theeconomicboosttotheGDPsofthesenationsisestimatedatUS$300$400billion,

Bymobilebroadbandwerefertonetworks,devices,andservicespoweredbycontemporaryhighspeedmobile datanetworks(principallyHSPA+andLTE). 2 UNCTAD,InformationEconomyReport2010atpages2425, http://www.unctad.org/Templates/webflyer.asp?docid=13912&intItemID=1397&lang=1.

1

5|P a g e

andrepresents1014millionjobs.InLatinAmerica,comparablefiguresareUS$5070billion andanadditional1.117millionjobs.3 Thus,thetrendsevidentinotherdevelopingcountriesaremanifestedinLatinAmericaaswell. Induecourse,ascharacterizedbyProfessorGeraldFaulhaberoftheWhartonSchool,[M]obile willbecometheportalofchoicetotheInternetformostoftheworldspopulation.4

NavigantEconomicsPresentationtoWirelessCommunicationsAssociationInternational,26February2010. GeraldFaulhaber,MobileTelephony:Economic&SocialImpact(4February2010), http://regulation2point0.org/wpcontent/plugins/downloadmonitor/download.php?id=25.

4 3

6|P a g e

3.SPECTRUMFORMOBILEBROADBAND 3.1.SPECTRUMDEMAND 3.1.1MARKETOVERVIEW

ThemarketintheAmericasformobilebroadbandisinrapidascent.Basedonthemostrecent subscription data from Informa, UMTSHSPA is the fastestgrowing mobile technology in the Americas(NorthAmerica,LatinAmericaandtheCaribbean),withsubscriptionsof116million at the end of 2010. Operators continue to deploy HSPA, the worlds most popular mobile broadbandtechnology,withatotalof77commercialnetworksin29countriesthroughoutthe region.5 The rate of growth, quarter over quarter, for UMTSHSPA at the end of December 2010 was 54%fortheUS&Canada(77milliontotalsubscriptions)and139%forLatinAmerica(38million total subscriptions). Both rates of growth surpassed the global quarteroverquarter rate for UMTSHSPA subscriptions of 40%. The combined quarter over quarter growth rate for the hemisphere at the end of December 2010 was 74%. These trends aredepicted inthe charts below.

4GAmericas,3GPPMobileBroadbandTechnologiesLeadtheWayintheAmericas,PressRelease(14March 2011),availableathttp://www.4gamericas.org/index.cfm?fuseaction=pressreleasedisplay&pressreleaseid=3076.
5

7|P a g e

Supporting this growth is the growing number of commercial HSPA networks in the region. Presently there are 13 commercially deployed UMTSHSPA networks in the U.S. and Canada, including5HSPA+networks.InLatinAmericaandtheCaribbean,thereare64HSPAnetworks commerciallydeployedin27countries,includingsevenHSPA+networks,asdepictedbelow.

LTEnetworksareemergingaswell.Inthesecondhalfof2010,VerizonWirelessandMetroPCS launchedLTEserviceintheUnitedStates.AT&TisexpectedtolaunchLTEmid2011.InLatin America, according to Maravedis Research, momentum is beginning to build as well. By the endof2011,Maravedisexpectstherewillbemorethan20LTEcommitments,anadditional7 LTEtrialswillbeconducted(inadditiontothe7thathavebeenconductedalready),andthere maybeupto4operatorswithcommerciallaunchesofthetechnology.Maravedisalsoexpects thatby20142015morethan25operatorswillhavecommercialLTEservices.6

MaravedisResearch,26January2011.

8|P a g e

3.1.2GROWTHINUSAGE

In its most recent Global Mobile Broadband Traffic Report, published in February 2011, Allot Communicationscalculatesthatmobiledatabandwidthusagegrewby73%inthesecondhalf of 2010, culminating in an overall annual growth rate of 190% for 2010. Over one third of globalmobiledatabandwidthconsistsofvideostreaming,whilefilesharingandwebbrowsing representanother30%and26%ofglobalmobilebandwidth,respectively.Thefollowingchart further illustrates these trends, including the clear upswing in video streaming over the past twoyears.

[Source:AllotCommunications,February2011]

TheAmericasisnotimmunefromtheimpactsofsubscriber,deviceandusagegrowthrateson mobilebroadbandnetworksbeingexhibitedinotherregionsoftheworld.Thesnapshotbelow of network key performance indicators (KPIs) from AsiaPacific (APAC), Europe and the Americas provides a glimpse of the accelerating levels of traffic being carried by mobile broadband networks from just before January 2007 (roughly a year following the first global launchofawidescaleHSDPAnetworkbyAT&T,thenCingular,during2005)toJanuary2010.

9|P a g e


[Source:NokiaSiemensNetworks,2010]

In its most recent The State of the Internet Report, for Q32010, Akamai provides additional insight on traffic growth in the hemisphere. Its analysis of consumption on mobile networks worldwide spans 111 providers, including 29 operators from 14 countries in the region (Argentina,Bolivia,Brazil,Canada,Chile,Colombia,ElSalvador,Guatemala,NetherlandAntilles, Nicaragua, Paraguay, United States, Uruguay, and Venezuela). Overall, data consumption on mobile networks grew quarter over quarter for 101 of the 111 operators, including five operatorsforwhichusagemorethandoubledquarteroverquarter.7 Lookingtowardsthefuture,IDATEestimatesthattotalmobiletrafficwillreachmorethan127 exabytes8(EB)globallyin2020,increaseover2010levelsbyafactorof33.
Totalmobiletraffic(EBperyearWorld) Europe Americas Asia Restoftheworld World 2010 1.03 0.78 1.65 0.41 3.86 2015 10.88 9.84 16.31 8.22 45.25 2020 28.15 27.33 43.85 28.48 127.82

[Source:IDATE,January2011]

7 8

Akamai,TheStateoftheInternet3rdQuarter,2010Report(January2011)atpage27. AnExabyteis1018bytes,orasRTTmemorablyexplains,equivalenttoallthewordseverspokenbymankind.

10|P a g e

By2020Asiawillrepresent34.3%oftotalworldmobiletraffic,Europe22%andtheAmericas 21.4%,accordingtoIDATEandasportrayedbelow.
Totalmobiletraffic(EBperyearWorld)
140.00 120.00 Yearly traffic in EB 100.00 Europe 80.00 60.00 40.00 20.00 2010 2015 2020 Americas Asia Rest of the world W orld

[Source:IDATE,January2011]

EricssonhasrecentlyestimatedsimilargrowthratesasIDATEformobiledataonaglobalbasis fortheupcoming5yearperiod.Bothfirmsforecastannualglobalmobiledatatrafficvolumes inexcessof40exabytesperyearby2015.Ericssonsprojectionisdepictedinthegraphbelow.

[Source:WirelessIntelligence,February2011]

11|P a g e

Ontheotherhand,AlcatelLucentforecastsevenhigherglobalmobiledatavolumesby2015.It predictsatwentyfoldgrowthinglobalmobiletrafficby2015to8,000petabytes(or8exabytes) per month, or 96 exabytes on an annualized basis.9 Alcatel Lucents forecast, as depicted below,isfueledinsignificantpartbyexpandingsmartphonedatagrowthovertheperiod.

[Source:AlcatelLucent,February2011]

The divergence in these figures is less important than their directionality and even more critically their drivers. Regarding the latter point, underneath these trends lie a variety of factors motivating a growing number of users to intensify and diversify their use of mobile broadband. The increase in both the number and capability of mobile broadband networks over the past couple of years as one driver is a given. Powerful new handheld devices and richer content and applications are additional factors. Finally, as connected devices increasingly penetrate the subscriber base, the number of devices per head will also add to thetotaltrafficonthenetwork.10 Considerationssuchasdevices,mobility,andrichapplicationsarenotsimplyadditive,butwork synergisticallyinawaythatcompoundstheimpactonmobilebroadbandnetworksmorethan
RTTsforecastofdatavolumeisonthehighsideaswell,predicting87exabytesofofferedmobiledatatraffic globallyby2015.SeeRTT,LTEUserEquipmentEfficiency&NetworkValue,atpage4. 10 CreditSuisse(6February2011)estimatesthatthenumberofdevicesperuniqueuserintheUnitedStateswill climbfrom1.2in2009to3.9in2015,asconsumersaddmobilebroadbandenabledlaptops,tabletsandconnected devicestotheirdevicecollections.
9

12|P a g e

justconsideringthesefactorsinamutuallyexclusivemanner.Forexample,Facebookcurrently reports that there are more than 200 million active users accessing the site through their mobiledevices,andthatpeopleusingFacebookontheirmobiledevicesaretwiceasactiveon Facebookasnonmobileusers.11 Moreover, high end devices continue to have distinct, multiplicative impacts. For example, Ciscos most recent Global Mobile Data Traffic Forecast Update (20102015), released 1 February2011,documentsthatrelativetononsmartphones,themultiplierforhighenddevice usage in 2010 as follows: smart phones (24x); handheld gaming console (60x); tablet (122x); mobile phone projector (300x); and laptop (515x).12 These multiplicative effects will persist through 2015 for all categories of devices, according to Cisco, as illustrated in the following chartlistingaveragetrafficperdevice(MB/month).
Device Type
Non Smartphone E-reader Smartphone Portable Gaming Console Tablet Laptop and netbook M2M module

2009
1.5 5 35 Not Available 28 1,145 3

2010
3.3 11 79 250 405 1,708 35

2015
54 245 1,272 879 2,311 6,522 166

[Source:CiscoVNIMobile,February2011]

Withrespecttowhatconsumersareusingmobilebroadbandcapacityfor,NorthAmericanand Latin American consumers exhibit important similarities and a few interesting differences. According to Sandvines most recent Global Internet Phenomena Report (Fall 2010), the top two mobile application categories in North America and Latin America during peak periods when bandwidth utilization is the heaviest are RealTime Entertainment (consisting of applicationsandprotocolspermittingondemandentertainmentconsumedasitarrives),and webbrowsing.Thiscontinuesthetrendinthedataseenin2H2009.Interestingly,whileP2P filesharingprofilesarefallinginbothregions,therelativeproportioninLatinAmericaisabout threetimestheproportioninNorthAmerica.Thefollowingchartsillustratethesetrends.

11 12

http://www.facebook.com/press/info.php?statistics CiscoVNI:GlobalMobileDataTrafficForecastUpdate(1February2011)atpage7.

13|P a g e


[Source:Sandvine,October2010]

Further expounding on these trends, TMobile USA CTO Neville Ray recently noted that the amount of data consumed by subscribers on TMobiles network is doubling every seven months, and that video use tripled on its network in 2010.13 Interestingly, the latter figure coincideswithGooglesrecentdisclosurethatYouTubenowexceeds200millionviewsadayon mobile,a3xincreasein2010.14AllotCommunicationsalsoilluminatesthistrendbyobserving that YouTube is responsible for an astounding 17% of overall global mobile data bandwidth. Thelattertwotrendsarefurthervisualizedbelow.

[Source:YouTube,January2011]

[Source:Allot,February2011]

Where are all these drivers ultimately headed? In short, an endstate exhibiting the ascendencyofthemobileinternet.By2012,accordingtoMorganStanley,moresmartphones

13 14

http://gigaom.com/broadband/tmobileshspadoublingdownonspeedsin2011/ http://youtubeglobal.blogspot.com/2011/01/musicvideosnowonyoutubeappfor.html

14|P a g e

willbeshippedthanPCs(includingnetbooksandnotebooks).Thisreflectsaninflectionpoint according to Morgan Stanley, suggesting the onset of a fundamental shift in how the typical consumerwillbeaccessingtheInternetinthecomingyears.

[Source:MorganStanley,November2010]

Providinganappropriatecodaforthisinflectionpoint,EricssonobservedattherecentMobile WorldCongressthatitexpectsmobilebroadbandsubscribershiptoreach1billionin2011,and potentially45billionby201516(outofatotalof78billiontotalglobalmobilesubscribers). Ericsson also anticipates that mobile data growth usage by 2015 will attain such levels as to comefullcirclewithtypicalusagelevelsonPCs.15

15

MorganStanley,Ericsson,QuickComment:FeedbackfromMWC,4February2011.

15|P a g e

3.2SPECTRUMREQUIREMENTS

Ifyoudon'tknowwhereyouaregoingyoumightwindupsomeplaceelse. YogiBerra Predictingthefuturecarriesobviousrisks,andpredictingfuturespectrumrequirementsisnot immune from this reality. Yet, the effort needs to be made while the future may be unknowable, nevertheless it can be shaped and influenced. The effort to quantify the incremental spectrum needed to support mobile broadbands continued vibrancy is a critical exercise,evenifitisanimpreciseone. Ataconceptuallevel,spectrumforecastsrequiredataandassumptionsinthefollowingareas: Available spectrum: current and foreseen levels of allocations for the relevant study period, tempered by the industry structure (e.g. operator market shares; incumbent versusentrantstatus,etc.) Traffic demand: typically parameterized in terms of peak or busy hour throughput required,andhighlysensitivetouserprofiles,devicesinuse,andservicesdemanded Networkdesign:assumptionsrelatedtothenetworkarchitecture,encompassingsuch factors as spectrum portfolio (e.g. high versus low bands), subscriber demographics (rural versus urban versus suburban), network topology (e.g. cell densities, degree of utilizationofoffloadingviafemtocellsand/orWiFi,etc.);radioaccesstechnology(e.g., EDGE,HSPA,HSPA+,LTE,etc.) NetworkCosts:estimationsofRANCAPEXandOPEXlevelsrequiredforagivennetwork design, including potential spectrum costs, which factors into investment considerations, that is, tradeoffs among various options to address future capacity needs(e.g.cellsplitsversus.additionalspectrumpurchases)

Spectrum forecasts are complicated and highly detailed undertakings, involving the developmentofcomplexdemand/capacitymodelsbasedontheabovefactors.Forpurposes of this paper, we have not undertaken our own forecast, but have chosen to survey and summarizethefindingsofrecentforecastsconductedataglobalornationallevel,andtodraw reasonableinferencesfromthiseffort.Thesummaryanddistillationofkeypointsfollows.
3.2.1.ITURM.2078(2006)

ITUR M.2078, published in 2006, is probably the most frequently cited long term spectrum forecastforthemobileindustry.TheITUundertooktodeterminehowmuchspectrumwould be needed per country in the years 2010, 2015 and 2020. The table below summarizes the resultsoftheITUsanalysis,whicharebrokendownbyhigherorlowermarketdevelopment

16|P a g e

status compared to a single global common market, as well as by Radio Access Technology Group(RATG).RATG1coverspreIMTandIMT,aswellasenhancementstoIMT,andRATG2is comprisedofIMTAdvanced.
Market setting Year Higher market setting Lower market setting Spectrum requirement for RATG 1 (MHz) 2010 840 760 2015 880 800 2020 880 800 Spectrum requirement for RATG 2 (MHz) 2010 0 0 2015 420 500 2020 840 480 Total spectrum requirement (MHz) 2010 840 760 2015 1 300 1 300 2020 1 720 1 280

[Source:ITURM.2078,2006]

The ITU concluded that total spectrum requirements for the higher market setting would be 840MHzby2010,1300MHzby2015and1720MHzbytheyear2020.Evenforthesituationin which a lower level of market development is assumed, the ITU projected total spectrum requirementsof760MHzby2010,1300MHzby2015and1280MHzby2020.16 Extrapolating from the ITUs forecast, in 2007 the NGMN Alliance (a global coalition of operators, industry partners, and academic advisors) determined what the net spectrum requirement would be, based on existing allocations, in each of the three ITU regions. The followingchartpresentstheNGMNAlliancesfindings,whichinsummaryarethatbetween500 MHzand1GHzwouldbeneededdependingonregionby2020.17

[Source:NGMNAlliance,June,2007]

EstimatedSpectrumBandwidthRequirementsfortheFutureDevelopmentofIMT2000andIMTAdvanced, ReportITURM.2078(2006). 17 SpectrumRequirementsfortheNextGenerationofMobileNetworks,NGMNAlliance(20Jun.2007)atp.22, availableat http://www.ngmn.org/uploads/media/Spectrum_Requirements_for_the_Next_Generation_of_Mobile_Networks. pdf.

16

17|P a g e

3.2.2NATIONALBROADBANDPLAN(2010)

In March 2010, the US Federal Communications Commission (FCC) released its National BroadbandPlan(NPB),areportmandatedthepreviousyearbytheUSCongress.Amongother matters, the NPB described the critical role that mobile operators will play in meeting US broadband objectives through the remainder of the decade, as well as the incremental spectrumrequirementthatmobileoperatorswillrequireinordertodoso.18 Inshort,theNPBconcludedthattheUSGovernmentshouldmake500MHzofnewspectrum availableforbroadbandwithinthenexttenyears,ofwhich300MHzshouldbemadeavailable forusewithinfiveyears.Further,pursuanttoanExecutiveMemorandumtofederalagencies issuedinJune2010,U.S.PresidentObamadirectedagenciestoworktosecurethisincremental 500MHzofspectruminatimelyfashion.19 TheFCCpublishedatechnicalpaper,MobileBroadband:theBenefitsofAdditionalSpectrum,20 inOctober2010,thegoalofwhichwastomakeareasonabledemonstrationthatmobiledata demandislikelytoexceedcapacityundercurrentspectrumavailabilityinthenearterm.The FCCsconclusionswerethreefold: Itislikelythatmobiledatademandwillexhaustspectrumresourcesinthenextfive years; Aspectrumdeficitapproaching300MHzislikelyby201421 Anarrowlycircumscribedestimationoftheeconomicbenefitfromreleasing additionalspectrum(thatistheavoidanceofunnecessarycostsinsatisfyingmobile datademand)islikelytoexceed$100billion(USD).

NationalBroadbandPlan,Chapter5:Spectrum(March2010),http://www.broadband.gov/plan/5 spectrum/#_edn67. 19 http://www.whitehouse.gov/thepressoffice/presidentialmemorandumunleashingwirelessbroadband revolution 20 http://download.broadband.gov/plan/thebroadbandavailabilitygapobitechnicalpaperno1.pdf 21 CreditSuissesMay2010analysisoftheNPBfurtherdemonstratestheloomingspectrumgap.UsingtheFCCs assumed rate of increase in data consumption and interpolating to 2015 (yielding a 72x increase in data consumption by 2015), 300 MHz incremental spectrum (1.6x increase over current levels), and accounting for improvementsinutilizationandspectralefficiency,65%offuturedemandwouldstillbeunmet.Itishard,Credit Suisse argues, to imagine that these capacity gains can be achieved by further cell splitting. Credit Suisse, TelecomIndustryThemes:ProfitingfromtheSpectrumCrisis,24May2010,atpages45.
18

18|P a g e

3.2.3.MOBILEBROADBANDCAPACITYCONSTRAINTS(RYSAVYRESEARCH)

In February 2010, Rysavy Research published Mobile Broadband Capacity Constraints & the Need for Optimization,22 a report which included results of a spectrum demand modeling exercise suggesting that many operators spectrum could be consumed within three to five years. The model is based on an examination of how much data users consume in a month, which depends in part on the type of device. It calculates the bitpersecond load per broadband subscriberperdevicetypeduringthebusiesttimesoftheday.Themodelthenmultipliesthe peruser traffic amount by the number of mobile broadband users in a typical cell sector to obtainatotaldataloadinthatsector.Then,knowingthespectralefficiencyofthetechnology beingused,themodeldeterminestheamountofspectrumneededtosupportthatload.The model makes certain projections to account for growth in mobile broadband usage and increasesinspectralefficiencyovertime,andcalculatestheamountofspectrumthatalarge operatorwouldneedtosupportthegivendemandortotaldataloadinasector. Thefollowingchartdepictstheamountofspectrumanoperatorwouldrequireinitsbusiest marketstomeetforecasteddemand.

[Source:RysavyResearch,February2010]

22

http://www.rysavy.com/Articles/2010_02_Rysavy_Mobile_Broadband_Capacity_Constraints.pdf

19|P a g e

Anotherwaytoconsiderthistrend,accordingtothereport,istocomputeaveragedatausage acrossallsubscribersandthencomparethatwiththeaveragecapacityforeachdatauser,using agenerousassumptionthatanoperatorhas50MHzofspectrum(25MHz+25MHz)deployed forjustbroadbanddataservices(1020MHzofspectrumforbroadbandismorecommonfor thetypicaloperatortoday).

[Source:RysavyResearch,February2010]

These results are broadly consistent with the FCCs conclusion in its October 2010 technical paper, noted above, that demand is likely to outstrip supply by 2015, if not earlier, in the absenceofincrementalspectrum.
3.2.4.SPECTRUMMARKETS:MOTIVATIONS,CHALLENGES,ANDIMPLICATIONS(2010)

ProfessorsBerry,HonigandVohraofNorthwesternUniversity,writingintheNovember2010 issue of IEEE Communications Magazine, outlined a vision of an expansive twotier spectrum market(consistingofasecondarymarketforownerstradingspectrumassetsandaspotmarket forlimiteddurationrentalsatparticularlocations).Afundamentalpremisetotheworkability ofsuchamarketis,accordingtotheauthors,theassumptionthatspectrumisindeedascarce resource. To test this hypothesis, they estimated the achievable rate per user assuming all spectrum assigned to nongovernment services between 150 MHz and 3 GHz in the United States is available for mobile broadband access. The total bandwidth is 1,018 MHz, and the specificspectrumbandsareprovidedinthefollowingtablefromthearticle.

20|P a g e


[Source:IEEECommunicationsMagazine,November2010]

Theauthorsdonotgosofarastoestablishspecificlevelsofscarcityovertime.However,their conclusionsareinstructive.Basedontheirmodeling,theyfindthatwithgenerousassumptions as to the availability of a managed infrastructure, coordinated frequency reuse, extensive spectrumsharing,relativelysmallcellradii,andothers,dataratesinexcessof1Mbpscouldin principlebemadecontinuouslyavailabletoeverymemberofadenseurbanpopulations. Evenso,theauthorsacknowledgethatthisrateisoptimisticasaresultoftheassumptions made.Asapracticalmatter,theyfind,aspectrumshortfallisarealriskintheUnitedStates, anddemandforspectrummayexceedsupplyasusers,applicationsandsystemsproliferate.
3.3TAKEAWAYS

Intheprecedingsectionwehaverevieweddatareflectingthedemandformobilebroadbandin theAmericas.Theregionispoisedforacceleratedgrowthindatabeingtransitedovermobile broadbandnetworks.Existingstudiesofwhetherthesenetworkswillhavesufficientcapacity tomeetburgeoningdemandarelimitedandnotsusceptibletooneononecomparison.Some importantbroadimplicationscan,however,bedrawn. 1.DEMANDISLIKELYTOOUTSTRIPSUPPLYINSHORTORDERUNDERABUSINESSAS USUALAPPROACH. Intheabsenceofsignificantactionstoaddresscapacitytrends,mobilenetworksarelikelyto begin experiencing capacity crunches in the next several years. It is unlikely, observes GoldmanSachs,thatthecurrentspectrumavailabilitywillbeabletosupporttheamountof datatrafficthatisexpectedtooccur.23CreditSuisserecentlyconcludedthat,withregardto

GoldmanSachsEquityResearch,LTE:fuelingthemobilesupercycle;implicationsacrossTMT(9February2011) atpage27.
23

21|P a g e

the United States, demand for wireless capacity will outstrip what can be offered by all spectrumcurrentlyavailabletowirelesscarriersby2013.24 While the situation may be less well quantified in the rest of the hemisphere, this is not a reasontodelayaddressingtheissueintherestoftheregion.Asexplainedfurtherbelow,other countries in the hemisphere generally begin with a relatively smaller spectrum portfolio for mobilebroadbandcomparedtotheUnitedStates.Whiledemandmayempiricallybesmallerin thesecountriescomparedtotheUnitedStates,itisalsotruethatnewspectrumallocationswill take considerable time to put to use, including potentially for spectrum whose utilization dependsonthebroadcastTVanalogswitchover. 2.THEREISNOSINGLEPANACEATOADDRESSINGTHISGAP. Historically,capacityonmobilebroadbandnetworkshasbeenaugmentedinthreebasicways (1)drivingincreasedspectralefficiencyfromspectrumallocations(asmeasuredinbitsperHz); (2) increasing the number of sites in the network so as to intensify frequency reuse; and (3) securingadditionalspectrum.25 Network operators have a long history, continuing up to the present, of driving spectral efficiency improvements and investing significant resources to densify their networks. Regarding the former, the evolution of 3GPP technologies from EDGE to HSPA to LTE is reflectedintheincreasinglevelsofspectralefficiencyovertime.26Thisisillustratedinthechart below.

CreditSuisseEquityResearch,Clearwire(6February2011)atpages2728.CreditSuisseforecaststhatwithout additionalspectrumdatacapacityshortfallintheUnitedStateswouldreach400Kterabytes(TB)permonthby 2015.(ATBis1012bytes.) Capacityissuescanalsobeaddressedfromthedemandside,thatis,byprovidingtoolsandincentivesforusers tobetterforecastneeds,includingusagebasedserviceofferingsaswellasgraphicalusertoolstohelpconsumers monitorandapportiondataconsumption.Thisisapotentiallyimportantcomponentofaddressingcapacity shortfalls,butliesbeyondthescopeofthepresentreport. 26 EnhancementstothespectralefficiencyofHSPA+andLTEcontinueunabated,exemplifiedbytheconsensus reachedatthemostrecent3GPPRANPlenaryrelatedtoadditionalfeaturesetsforHSPA+beyondRelease10.See DraftReportofDecember20103GPPRANPlenary, http://www.3gpp.org/ftp/tsg_ran/TSG_RAN/TSGR_50/Report/Draft_Report_RAN_50_friday.zip;4GAmericas,4G MobileBroadbandEvolution3GPPRelease10andBeyond,Section7(February2011).
25 24

22|P a g e

[Source:CreditSuisse,February2011]

Similarly,topologiesformobilenetworkshavebeencharacterizedbyanincreasingnumberof smallersizedcells.Thetrendtowardmoreandmorecellswithdiminishingcellradiiisforecast tocontinueinthenearterm;AlcatelLucentprovidesoneestimatebelow.

[Source:AlcatelLucent,June2010,(basedonYankeeGroup2008analysis)]

Cell densification and spectral efficiency gains are bounded, however, by basic technical and financial realities. Incremental site builds are frequently time intensive, involving additional zoningand/orleasingapprovals.Byincreasingthenumberofcelledges,operatorsalsoriskan increase in the potential for intersite interference, adding to the complexity and resources neededforRFnetworkdesignandmanagement.Finally,densificationissubjecttodiminishing returns on invested capital as the number of end users that can be served in a financially prudentwayshrinkspercellsiteastheresultofmoreintensivereuse. Similarly, the spectral efficiency gains that can be accrued from technology innovation also comewithaprice.Thesegainsmaybeaccompaniedbyincreasedcomplexity(andattendant

23|P a g e

costs) in end user devices and network equipment. In addition, technology innovation is subject to diminishing returns in the form of the Shannon Bound. The Shannon Bound is a theoreticallimittotheinformationtransferrate[perunitbandwidth]thatcanbesupportedby agivencommunicationslink.TheboundisafunctionoftheSignaltoNoiseRatio(SNR)ofthe communications link. As shown in the figure below, current mobile broadband technologies areallwithin2to3decibels(dB)oftheShannonBound,indicatingthatthereislittleroomfor improvementinspectralefficiency,speakingstrictlyfromtheperspectiveofinnovationsinthe radioaccesstechnology.

[Source:4GAmericas,September2010]

The technical and economic limits on cell densification and spectral efficiency as a means to bolster capacity mean that each technique cannot be relied upon individually to address capacity concerns. In fact, there is no single panacea to address future mobile broadband capacity requirements. Both methods will continue to be vital, as will be the importance of additionalspectrumallocations,thesubjectofthenextsection.27

Additionalmeasurestoaddressmobilebroadbandcapacityexist,andthesealsoneedtobethoroughly investigatedandadoptedwhereappropriateforagivenoperator.Thesetechniquesincludebutarenotlimitedto: (a)WiFioffloading;(b)femtocells;and(c)applicationlayercompression.

27

24|P a g e

3.SUPPLEMENTALSPECTRUMALLOCATIONSAREACRITICALPARTOFADDRESSING THISGAP. Eachcountryfacesuniquecircumstancesintermsofmobilebroadbanddemandandavailable spectrum.Thecommonthreadunitingall,however,istheloominggapbetweenthetwo.28The Broadband Commission for Digital Development, established last year by the United Nations and UNESCO, perhaps summed it up best in its September 2010 report, A 2010 Leadership Imperative The Future Built on Broadband, when it stated that countries must radically rethinktheavailabilityofadequateradiofrequencyspectruminthebroadbandera.29 TheUnitedStatesmaybeanoutlierinthehemisphereintermsofthequantityofspectrum allocatedtomobilebroadbandincomparisonwithothercountries,assuggestedinthe followingchartfromCTIAtheWirelessAssociation.

[Source:CTIA,May2010]

Note that the chart above does not reflect AWS auctions concluded in Mexico (as well as Germany)lastyear,orpreparationsbeingmadeinCanadaandMexicotoauctionspectrumat 700MHzandelsewhereasspectruminthepipeline.

Althoughthetopicofthispaperonmobilebroadbandspectrum,growingdatausageparticularlyfrom smartphonestakesitstollonotherpartsofthenetworkaswell.Smartphonesarenotoriouslychattyorinother wordsdemandinordinatesignalingresourcesfromthenetwork.Thisintensifiesdemandforspectrum,butin additionhascapacityimplicationsfornetworknodes(especiallyRadioNetworkControllers(RNCs)andServing GPRSSupportNodes(SGSNs))thatmustprocessthesesignals.Itmayoftenbethecasethatthesenodeswere notdimensionedforthesmartphoneonslaughtwhentheywereinstalled,meaningthattheymustbeupgradedin ordertohandleescalatingsignalingloads.RTTestimatesbasedonpresenttrendsthatsignalingvolumeinmobile networkswillhaveincreased40%between2008and2012.RTTEconomics,LTEUserEquipment,Network Efficiency&Value(September2010),atpage45. 29 TheBroadbandCommissionforDigitalDevelopmentDeclarationandReport(19September2010)atpage6 (emphasisadded),http://www.broadbandcommission.org/report1/report1.pdf.
28

25|P a g e

Notwithstanding,allcountriesinthehemispherewillconfrontacriticaldecisional juncturein the short term. The NGMN Alliance, as described above, extrapolated from ITUs 2006 spectrum forecast to determine that 500 MHz 1 GHz of incremental spectrum would be neededbycountriesintheAmericasby2020,dependingonthelevelofmarketdevelopment. Observingthatcongestion onbroadbandnetworkissettoworsen, ITUSecretaryGeneralDr. HamadounTourrecentlycalledongovernmentstotakeurgentactionnowtosupportmobile broadband growth, remarking that steps including greater spectrum availability are imperativeinordertoavoidnetworkbottlenecks.30Willcountriesheedthismobilebroadband calltoaction?

ITU,NetworkcongestionsettoworsenITUcallsforinternationalbroadbandcommitment,11February2011, http://www.itu.int/net/pressoffice/press_releases/2011/01.aspx.
30

26|P a g e

4.SPECTRUMANIMATINGPRINCIPLES 4.1WELLCONCEIVEDSPECTRUMSTRATEGIESAREVITAL

In this section, we describe a set of principles that 4G Americas believes should animate the developmentofstrategiesforensuringthecontinuedvitalityofmobilebroadbandservicesin theAmericas.Wepresentprinciplesintheareaofspectrumallocation,spectrumassignment, andtiming.Togetherwiththeseprinciples,wepresentspecificillustrationsofhowindividual countries have pursued or are pursuing efforts which are generally aligned with these principles. By instantiating these principles, meaningful guidance can be provided on how to bringthemtolife,toconvertwordsintoaction.
4.1.1CORESPECTRUMALLOCATIONTENETS

Notwithstanding continuing innovation and investment that has led to more intensive use of spectrum assets, suitable and sufficient additional spectrum is critical to meet the growing demandformobilebroadband.4GAmericasbelievesthatthefollowingtenetsshouldanimate efforts to allocate spectrum resources. The principles outlined below will further stimulate investments,keepcostsdown,andultimatelyhelpmeetdemandforbroadbandservicesina timelymanner. CONFIGURELICENSESWITHWIDERBANDWIDTHS Technologies that support ever more supple mobile broadband capability, such as LTE, will increasingly require wider bandwidth channels to meet consumer demand for bandwidth intensive and contentrich services.31 Wider spectrum blocks maximize spectrum use by accommodatingmorebitsandallowingmoreresourcestobepooledforsharingamongusers. Spectrum allocations should therefore be in sufficiently large, contiguous blocks to accommodate the future development of mobile broadband networks. Particularly in urban andsuburbanareas,allocationsshouldfocus,ataminimum,on2x10MHzblocks.

HSPA+performancecanbeenhancedbycombiningtwoormore5MHzcarriersinatechniqueknownascarrier aggregation.LTEcanbedeployedinarangeofbandwidthsfrom1.4,3,5,10,15to20MHz,andcanalsobenefit fromcarrieraggregationtechniques.Fromaspectralefficiencystandpoint,however,carrieraggregationcannot beconsideredasubstituteforwiderchannelsinthefirstinstance.

31

27|P a g e

 GROUPLIKESERVICESTOGETHER Current spectrum allocation frameworks tend largely to compartmentalize spectrum by types of services. The grouping of like services can reduce complexity and cost, and allow more flexibility in the form factor of the subscriber equipment. For example, allocating additional spectrumadjacenttosimilarservicesandwithsimilarduplexdistancesreducesthenumberof bands that a device must support. Standardsdevelopment would as a result beaccelerated, and existing user equipment would be more easily modified rather than requiring new technologydevelopments,whichacceleratesthemarketavailabilityofthosedevices. This would be accomplished because the selection of the additional spectrum would not increasethenumberofbandstobesupportedinamultibanddevice.Weofferaspotentially positive developments in this area two inquiries recently begun by Mexico and the United States to consider allocating additional spectrum for mobile broadband above 1755 MHz and 2155 MHz. These efforts are in alignment with 3GPP Band 10 (the extended AWS Band directly adjacent to 3GPP Band 4), which is harmonized throughout the Americas. 32 Supplementingcurrentallocationsinthiswaywouldreducedevicecomplexityandcostthe additional spectrum band can be implemented largely as an extension to an existing band. Designatingspectrumforcertainservicesadjacenttolikeservicespectrumbands,withsimilar duplexdistances,wouldconsequentlymaximizeperformanceandefficiencies. BEMINDFULOFGLOBALSTANDARDS. Technical standards are the foundation for service providers and manufacturers to develop competitive products and services. Further, standards help facilitate regulatory compliance, establish patent policies for use of essential technologies, provide a platform for third party supplier solutions, and serve to energize investment. Finally, standards allow companies to take advantage of economies of scale that lower costs and promote growth, maximizing opportunitiesforinnovation. Inmanyinstances,globallyacceptedstandardsaredevelopedwithspecificspectrumbandsin mind, and take into consideration coexistence with adjacent services. Therefore, whether certain spectrum bands have been harmonized and whether a standard exists should be factoredintospectrumallocationdecisionmaking.
Seehttp://www.ntia.doc.gov/press/2011/500mhzstatement_02012011.html(NTIAannounces31January2011 thatitwillbegindetailedanalysisofwhethergovernmentspectrumat17551850MHzcanberepurposedfor commercialbroadbanduse);http://www.cft.gob.mx/es/CofetelCofetel_2008/Reporte_Consulta_Publica_Espectro (Cofetelsummaryissued8February2010ofpublicconsultationinitiatedinNovember2010onseveralspectrum bands,including175570MHz&215570MHznotably,allrespondentsexpressedinterestinthesebands).
32

28|P a g e

 PURSUEHARMONIZED/CONTIGUOUSSPECTRUMALLOCATIONS. Ensuring that spectrum allocations are, to the greatest extent possible, in accord with internationalallocationspromotesinnovationandinvestmentbycreatingcriticaleconomiesof scale.Similarly,harmonizationfacilitatesglobalroamingandhelpscountriesthatshareborders managecrossborderinterference.Withoutharmonization,newtechnologiesandservicescan be difficult to export to other markets, and internal markets will fail to benefit from developments in international markets. Fragmented spectrum allocations hamper innovation and require companies to dedicate resources to develop new or adapt existing products or technologiesforasinglemarket,ratherthansharingthosedevelopmentcostsglobally.This,in turn, increases the costs and limits the potential availability of products and services for the consumer. Moreover, new/modified technology takes time to commercialize, delaying the availabilityofservicesanddevicesinthemarkettomeetburgeoningmobiledatademand. OneofthemainreasonsOFDMwasselectedasthemultiplexingtechniqueforLTErelatestoits abilitytosubstantiallyincreasespectralefficiencywhenthetransmittedRFsignaliscomposed of a relatively large number of orthogonal subcarriers. The latter requires a relatively large amountofcontiguousspectrum.Theresultingincreaseinspectralefficiencyisinthenatureof astatisticalgain,alsoknownastrunkingefficiency.Trunkingefficiencyincreasesasdatafor various users are scheduled (multiplexed) for transmission over larger numbers of sub carriers,anditsmagnitudeisthusdeterminedbytwofactors:(a)therobustnessofthenetwork scheduling algorithms, and (b) the amount of spectrum the RF signal occupies. From an operationalstandpoint,OFDMbasedtechnologieslikeLTEcanexhibitasignificantincreasein spectralefficiencyiftheoccupiedspectrumisatleast10MHz(or10x10MHzifconsideredon apairedbasis). EXHAUSTEXCLUSIVEUSEOPTIONSBEFOREPURSUINGSHAREDUSE. The exclusive use model refers to a framework in which a licensee has rights that are exclusive, flexible, and transferable, while at the same time inheriting specific responsibilities thatcomewiththeserights.Inanexclusiverightsmodel,thelicenseeinsomecasescanlease some or all of the spectrum usage rights associated with their licenses to third parties. Licenseesarealsoaffordedprotectionfrominterferenceandmayusethespectruminaflexible manner consistent with the terms of the license. Moreover, the certainty provided by the exclusiveuselicensingapproachencouragesinvestmentbypromotinganenvironmentinwhich interferenceandsystemcapacitycanbepredicted. Successfulsharingarrangementsmustbetailoredtothespecificconditionsintheband,ona bandbybandbasis.Theseconditionsmustbeknownatthepointofdesigninceptiontoavoid inappropriate and costly technology development. Because each sharing environment is

29|P a g e

unique, technology research and development suitable for the services sharing frequency bands and knowledge of the specific sharing constraints are required; otherwise, investment andinnovationwillbeimpededbysuchoperationaluncertainties. Somesharingenvironmentsarelessconstrainedthanothers.Theseenvironmentsarecreated bygroupinglikeserviceswheretechnologiesandtechnicalrulesaresimilaramongoperators. Forexample,afixedservicecanbeeasiertopredictandengineerthanamobileservice,and with appropriate licensing or notification requirements, can provide regulatory certainty a prerequisiteforinvestment. Inaspectrumsharingenvironment,spectralresourcesaremadeavailablebasedonestablished technical etiquettes or standards that set power limits and other criteria for operation of devicestomitigatepotentialinterference.Inaddition,thesharingenvironmentmustbewell understood sharing criteria may include geographic protection zones, power limits and temporalrestrictions.Forexample,spectrummaynotbeavailableatcertaintimesoftheday, ormaynotbeusednearcertainsites,forinstancenearradarsites. Theuncertaintyofspectrumaccessandpotentialsharingconstraintscantranslateintohigher network costs, and can limit the types of services that can be supported. Thus, proposals for spectrum sharing need to be closely examined, weighing the benefits and opportunities provided with a variety of other factors including service viability, technology support, and adequateknowledgeoftheoperatingenvironment. NOTALLSPECTRUMISFUNGIBLEALIGNALLOCATIONWITHDEMAND. Futurenetworkswillinalllikelihoodbenetworksofnetworksconsistingofmultipleaccess technologies, multiple bands, widely varying coverage areas, all selforganized and self optimized.33 Part of the consideration to support the development of such networks and theevolutionofdataserviceswillbetheavailabilityofsuitablespectrum,cellsitespacing, and increases in spectral efficiency to support an increase in the number of users and an increaseinthedatathroughputavailabletoeachuser. Essentially, these networks will need to address coverage requirements, that is, adequate serviceacrossabroadlydefinedgeographicarea.Thiswouldincluderuralandisolatedareas where the population density is low. In this case, the ability to provide services more efficiently is enhanced using spectrum with appropriate propagation characteristics, which arecharacteristicofthelowerradiofrequencies.Therearealsospecificneedsforcapacity,

33

Transitionto4G:3GPPBroadbandEvolutiontoIMTAdvanced,RysavyResearch/4GAmericas,Sept2010, http://www.4gamericas.org/documents/3G_Americas_RysavyResearch_HSPALTE_Advanced_FINALv1.pdf.

30|P a g e

in which greater bandwidth is needed across a smaller area, but by a number of interests contending for limited spectral resources. This tends to be the case particularly in dense, urban environments. Even with the latest digital air interfaces and other new network features, carriers still need to add new cell sites or purchase additional spectrum to supplementnetworkcapacity.34 Stated differently, when considering frequency bands and combinations thereof, device manufacturers often categorize bands in terms of high band or low band designations. High bands are frequency bands above 1 GHz (for example, 1900, 2100, and 2100 MHz); low bands are the frequency bands below 1 GHz (examples 700, 850, and 900 MHz). The higher bandsaregenerallyeasiertoimplementinadeviceintermsofsize, asthehigherfrequency bands require shorter antenna length. The lower bands require longer antennas for optimal performance. In 3G and 4G devices with multiband support, the number of high bands supportedalwaysexceedsthenumberoflowbands.SeetheaccompanyingchartforUShigh andlowbandallocations.

[Source:CreditSuisse,May2010]

InFocus:TheCapacityChallenge,RossErnst,TenXcWireless, http://connectedplanetonline.com/wireless/technology/infocus_capacity_challenge_112706
31|P a g e

34

Propagation characteristics are sometimes overlooked, but their impact can be tangible and very significant for overall design and performance. Suitable lower band frequencies can penetratewallsofbuildingsandhaveasignificantcoveragerangewithoutrequiringthemobile handsettosupportanunwieldyantenna. FutureIMTAnetworkswillsupport100Mbpsforhighmobilityand1Gbpsforlowmobility.35 The spectrum to serve those areas where the demand for services reaches these levels will largelycomefromhigherfrequencybandsthathaveadequatecontiguousspectrumtosupport thetrafficgenerated. The implementation of future networks must support access to communications anywhere, anytime. Therefore these networks will rely on multipleaccess technologies supported by a variety of spectrum bands to address both coverage and capacity needs. This will require suitablelowerfrequenciestosupportinbuildingcoverageaswellassparselypopulatedareas, inadditiontohigherbandssupportingcapacityobjectivesthroughamixofmacro,microand picocelldeployments.

ITURM.1645,FrameworkandOverallObjectivesoftheFutureDevelopmentofIMT2000and SystemsBeyondIMT2000(2003).

35

32|P a g e

4.1.2SIDEBARWHYDOESALLTHISMATTER?

WhyDoesAllThisMatter? TheImpactonDevices,Equipment,CostandPerformance
A.SpectrumBandProliferation/Fragmentation Ifonegoesback1015yearsintime,inthecontextofGSM,devicemanufacturersonlyhadto consideroneortwofrequencybandswhendesigningamobilephone.MultibandGSMdevices which allowed for global roaming soon followed in the early 2000s, bringing support for 23 frequencybands,typicallyincludingabandsolelyforglobalroaming.Withthemid2000scame the advent of UMTS/WCDMA technology, creating the need to support local 3G frequency bandsinadeviceinadditiontotherequiredGSMbandsforbothlocalandglobalroaminguse. Notsurprisingly,thedemandforglobalroamingwith3Gputfurtherpressureforinclusionof additionalroamingbandsinhandsets. The advent of LongTerm Evolution (LTE) technology will place additional frequency band requirements on mobile device manufacturers, as LTE supports interworking with the legacy technologies. As LTE technology matures and home network coverage increases, device manufacturers will beexpected to support the local and roaming bands which support this technologyaswell.This,inpractice,couldmeansupportingtwotothreeadditionalfrequency bands, on top of what is required for the legacy technologies. Currently, over 30 frequency bandsaresupportedinthestandardsforLTE,asdepictedbelow.

33|P a g e

AsnetworkcapabilitiesevolvefurtherintotheHSPA+domainandentertheLTEdomain,smart phones and their associated features have become a growing area of industry attention and benchmarking. One common benchmark of a mobile device identifies the radio technology supportedinaparticularfrequencybandsupport(seeoneexamplebelow).

[Source:www.tmobile.com]

Onabasiclevel,thisprovidesanideaofwhere(oronwhatnetwork)thedevicecanoperate andwhatlevelofoperationcanbesupported(e.g.GSM/EDGEversusUMTS/HSPA). Today,mostsmartphonessupportthefourGSM/EDGEfrequencybands,allowingforsupport ofbothlocaloperationsandglobalroaming.ThelocalGSM/EDGEoperationsintheAmericas includemainlythe850and1900MHzbands,andtheGSM/EDGEglobalroamingbandsinclude the 900 and 1800 MHz bands. When considering the 3G band, the focus in the Americas is typicallyon23frequencybands.Theseincludethe850MHz(cellular),1900MHz(PCS)andthe AWS(1700/2100MHz)band,dependingontheoperatorsrequirements.Incaseswhereglobal 3Groamingsupportisdesired,the1900/2100MHzbandprevails,withthe900MHzbandbeing concurrentlysupportedinsomecases. Consequently, WCDMA/HSPA implementations across the globe have resulted in spectral fragmentation into the aforementioned 5 major bands (850, 900, 1700/2100, 1900 and 2100 MHz bands). In the extreme case, a mobile device capable of supporting the 4 GSM/EDGE bandsand5WCDMA/HSPAbandswouldbecapableofsupportingfull2Gand3Goperationon agloballevel.Therearesomehighendsmartphonedevicesavailablewhichcansupportthis configurationtoday,butthislevelofsupportiscertainlynotwidespreadatthistime. Compoundingmatters,devicemanufacturersalsofacetherealitythatnotallfrequencybands are equal from an implementation standpoint. Considerations here include whether or not thebandisahighbandoralowbandaswellasthespectralproximityofagivenbandto otherbands.

34|P a g e

B.LimitsonFrequencyBandSupport In theory, there is no hard upper limit on the number of frequency bands that can be supported in a mobile device. However, there are very pragmatic limits on the number of bands, driven generally by cost, size, and performance issues. An additional frequency band increases the number of RF components required in the device, resulting in bill of material increasesaswellasadditionalpressuresonlimitedrealestateinsidethedevice. Theimpactofaddinganadditionalbandalsodependsonwhatbandisbeingaddedrelativeto whatotherbandsarepresentinthedevice.Iftheadditionalbandliesclosetoanexistingband, itmaybepossibletocoverthisadditionalbandwithsomeofthesameRFcomponentsandthe existing antenna(s). If the additional band is far away from other bands there is a much greaterimpactoncostandsize.New,possiblydedicatedRFcomponentsmustbeaddedandin somecasesevenanadditionalantenna(orantennas). Sometimes,antennasaredesignedtocovermultiplebandsinthemobiledevicewherefeasible. There can be a performance tradeoff in attempting to increase antenna bandwidth to accommodate new bands. These tradeoffs can be particularly acute if the new band is a prioritybandfortheoperator,whichmightotherwisedrivedevicevendorstobuildnarrower antenna bandwidths so as to obtain better overtheair performance. In this case, the cost versusperformancecalculationcanbeparticularlyvexingforoperatorsandtheirvendors. C.ImpactonDeviceComponents (1)BasicHandsetDesign. Atitsmostelemental,mobileRFdevicesconsistofoneormorefrontendmodules(FEMs) connectedtodigitalbasebandunitsaswellasoneormoreantennas.Onerepresentationofa multibandradiodesignidentifyingthesevariouscomponentsisprovidedbelow.

35|P a g e

[Source:RTTEconomics,September2010]

The FEMs manage the RF signal paths into and out of the device, and include numerous componentssuchasfiltersandamplifiers.Manyfrontendfunctionsareanaloginnature,as befitstheradiowavesexitingandenteringthedevice.Incontrast,thebasebandunitsperform thecomplexdigitalsignalprocessingrequiredtoencodeordecodeRFtransmissionsaccording to the technology embedded in baseband chips. Although baseband, FEM and antenna performance are closely interrelated, spectrum fragmentation has comparatively more significanceonthedesignandperformanceofantennasandFEMsthanbasebandunits,duein nosmallmeasuretotheirdependenceonthephysicallayoutofthevariousdeviceelements. (2)Antennas. A critical consequence of spectrum fragmentation relates to the feasibility of manufacturing deviceswithincreasinglymoresophisticatedantennaetechnology.Forexample,theabilityfor mobiledevicesusingHSPA+andLTEtechnologiestoprovideeverhigherdataratesovertime dependsontheincorporationofinnovativenewtechniques.Oncesuchinnovationisknownas MIMOMultipleInput/MultipleOutputwhichallowsthedevicetosenddatatoandfrom thenetworkonparalleldatastreams,thusboostingthetheoreticaldataratesofthesedevices. Asadesignmatter,MIMOrequiresthatmultipleradiopathsbesupportedwithinthespectrum bands designated for HSPA+ or LTE. In practice, this means multiple antennas must be

36|P a g e

implemented in the mobile device for these bands, meaning not only additional antenna elementsbutalsofrontendcomponents. Multiple antenna techniques such as MIMO depend on antenna separation, polarization and radiation pattern in order to achieve the best possible performance. If there is coupling betweentheantennas,thepositiveperformanceeffectofMIMOmightbereduced,particularly duetoincreasedlosses(i.e.,antennaefficiency).Theresultissuboptimalperformance,with datarateslowerthandesired,includingthepossibilityofbeinglowerthaninthesingleantenna case. Asaveryroughestimate,aminimum0.15wavelengthseparationbetweenantennasisneeded toallowamobiledevicetoachieveitstargetperformancelevels.Inthecaseof700MHzthis wouldmeanaphysicalseparationofapproximately2.5inchesbetweentheantennas.When considering the 1900 MHz band this would translate to roughly 1 inch. Clearly, there is a sizeable difference between these separation requirements when considering support for differentspectrumbandsandtheconsequentimpactonthesizeofthedevice. Experience from commercial launch LTE networks suggests that MIMO can perform in small mobile device form factors comparably in lower frequencies as in higher frequencies, notwithstanding the need for increased wavelength separations. The problem arises when more and more spectrum bands need to be introduced into the device to provide sufficient bandwidth(aswellasforotherconsiderationsincludingroamingandcoveragerequirements), whichpresentsadditional,significantdesignchallengesformobiledevices. Device manufacturers must also consider the following factors in their device designs when incorporatingmultipleantennatechnologies:

Frequencybandofoperation Signaltonoise(SNR)targets Interferencemitigationtargets(fromusersbothinsidethecellandinadjacentcells) Formfactorofthedevice OtherradiosinthedevicewhichmayincludeWiFi,GPS,Bluetooth,andFMRadio

Whenthesefactorsareconsideredandthenextendedtoincludemultiplefrequencybandsof operation,thedesignequationbecomesdramaticallymorecomplex.Theaboveconsiderations alldependtovariousdegreesonthefrequencybandsemployedinthedevice.Theotherradios presentinadevicemayoperatenearoneoftherequiredmobilefrequencybands,forexample, ormayrequiresomeadditionalisolationinsideofthedevicefromotherbands.

37|P a g e

(3)OtherComponents. Asthebandrequirementsformobiledevicescontinuetogrow,andbandsbecomeincreasingly fragmented, device manufacturers face growing challenges associated with finding the additionalrealestatewithinthedeviceforadditionalfrontendcomponentsandantennas.The additionalfrontelementsdonotscaletoquitethesamedegreeasbasebandunits,duepartly totheirdependenceonphysicallayoutandspaceconstraints. As a result, including new bands in a device adds nonrecurring engineering (NRE) and production costs. This would not prove problematic to the extent the market for the newly designed device was sufficiently large so that the costs could be amortized over a large customerbase.However,bydefinition,devicesincorporatingnonharmonizedspectrumbands will find appeal in only isolated markets, which save for China or India, will almost invariably lackthenecessaryscale. D.AdditionalBandProliferation/FragmentationImpacts Itshouldalsonotbeoverlookedthatspectrumbandproliferationandfragmentationcanhave impactsonthecostandperformanceofradionetworkinfrastructure.Existingradiosatbase stationsmayneedtobeupgraded,orentirelynewradiosmayneedtobedeployed.Increased CAPEXwillberequiredforthenewinfrastructure,whichfrequentlyalsoentailsincreasedOPEX for servicing and maintaining the equipment. See below for a schematic of modern base station architecture, wherein radios (remote radio units or RRUs) are increasingly situated closertotheantennaunits(AUs).

[Source:DesignArtNetworks(2010)]

38|P a g e

4.1.3ILLUSTRATIONS

Mexico In November 2010, Mexico initiated a consultation to solicit information on the potential to make spectrum available in three bands: 700 MHz, 1.7/2.1 GHz and 3.43.7 GHz.36 With respecttothesecondcategory,Cofetelsoughtspecificresponsestoquestionsabout175570 MHzand21552170MHz.This30MHzisan extensiontothe90MHzofAWS1frequencies which were auctioned in 2010. Together, they constitute 3GPP Band X, a band harmonized throughoutthishemisphereforHSPAandLTE.Thequestionsweredesignedtoprovideinput on interest in and optimal uses for the band, device ecosystem and infrastructure readiness, preferredchannelization,andpossibleauctionconcepts.37 Canada InJune2010,IndustryCanadacommencedaconsultationonthetransitiontoBroadbandRadio Service (BRS) in the 25002690 MHz band and associated modifications to the existing band plan.Withrespecttothebandplan,IndustryCanadaoutlinesthebenefitsanddisadvantages oftwopossibleoptions,oneconsistentwiththeUSallocationandtheotherconsistentwithITU Option1andwhichisbeingadoptedbynumerouscountriesacrosstheglobe.IndustryCanada concludesbyproposingtoadoptITUOption1andseekinginputontheimplicationsofsucha bandplan.38 Brazil Anatel is moving to accelerate theprocess of licensing the2.5 GHz band, which was recently reserved for mobile telephony service and aligned with ITU Option 1. According to the draft schedule,theauctionmayoccurinearly2011,atleastoneyearaheadofpreviousforecasts.If thetendertakesplaceinFebruary2011,LTEnetworksmaybedeployedinBrazilby2013.
http://www.cft.gob.mx/es/CofetelCofetel_2008/Cuestionario_para_la_banda_1721_GHz(11November2010). 4GAmericasfiledcommentsinsupportoftakingactionconsistentwiththeBandXregionalharmonization scheme. 37 Asnotedearlierinthispaper,theUnitedStateshasrecentlybegunanexaminationintowhethergovernment spectrumat17551850MHzcanberepurposedforcommercialwirelessbroadbanduse.Suchactionscouldalso besynergisticwiththeBandXallocationinthehemisphere. 38 ICconsultation:http://www.ic.gc.ca/eic/site/smtgst.nsf/eng/sf09881.html.4GAmericasfiledcommentsinthe proceeding,availableathttp://www.ic.gc.ca/eic/site/smtgst.nsf/vwapj/dgso001103GAmericas comments.pdf/$file/dgso001103GAmericascomments.pdf
36

39|P a g e

4.2MARKETORIENTEDASSIGNMENTAPPROACHESWORKSPECTRUMCAPSSHOULDBE DISFAVORED.

Since 1998, spectrum caps the total amount of spectrum that an operator can own have beenappliedbyLatinAmericanregulatorsinanattempttostimulatecompetitionbylimiting theabilityoftheestablishedoperatorstoacquirenewspectrum,fuelingnewentrantstoenter themarket,anddiscouragingmergers.Themajorityofcountriesintheregionhavethreeor four mobile operators competing in the market. With mobile penetration in Latin America above 96%, the region is ripe for the advancement of mobile broadband services. However, such deployment is being constrained by spectrum caps, which creates lengthy delays every time new auctions are planned. Existing operators with costefficient access to capital and operational knowledge are often prohibited from participating. Given the degree of competition in countries applying them, caps may not be necessary to ensure a competitive market, and may actually undermine those countries goals of deploying spectrumintensive mobilebroadbandtechnologies.Capsmaydiscourageoperatorinvestmentinbothurbanand ruralareas,andraisecostsforconsumers. Recentacademicworksupportstheseconclusions.BerkeleyUniversityProfessorMichaelKatz summarizestherisksasfollows: Spectrum caps act as a tax on success and can distort and limit competition. A binding spectrum cap can increase the costs of expansion for a service provider that has developed a successful business model that requires additional spectrum to meet consumer demand for its services. A spectrum cap therefore punishessuccessand,thus,discouragesfirmsfromcompetingto attractconsumersthroughimprovedservicesandlowerprices.39 Econometric modeling of the impact of spectrum caps in Chile, Argentina and Colombia suggests this very outcome caps can increase prices and lower service quality, the very objectivespresumablydisfavoredbypolicymakers.FranciscoMarroqunUniversityProfessor Wayne Leightons quantification of the impact of spectrum caps in these markets led to the followingconclusions:(a)halvingthespectrumavailabletoanoperatorforLTEfrom40MHzto 20MHzresultedinadoublingofcoststoprovideservice;and(b)afurtherhalvingto10MHz

DeclarationofMichaelL.Katz,PublicPolicyPrinciplesforPromotingEfficientWirelessInnovationand Investment,inFCCDocket0966,30September2009.
39

40|P a g e

led to a further quadrupling of the cost of LTE service. Higher costs invariably lead to lower demand,andhenceloweradoption,ofmobilebroadband.40 Asthemobilewirelessindustrymovestothenextgenerationoftechnologiesandservices,the spectrumcapsofthe1990sseemantiquatedandbuiltuponapasteraoflowpenetrationvoice services.Inmanycountriestheoldspectrumcaprulesarebecominglimitingfactorstobringing tomarketnewbroadbandservicesandapplicationsthatwouldbenefitsociety. As spectrum regulators review their countries specific caps, they should consider a few key questions: 1. Does significant and measurable operator competition exist in the marketplace? 2. Is enough spectrum available today for operators to commercially deploy mobile broadband data technologies such as HSPA+ and LTE in the next few years? 3. Is the country proactively aligning with International Telecommunication Union(ITU)orotherestablishedguidelinesfortheamountofspectrumrequired tofomentthenextgenerationofwirelesstechnologies? The following table shows the current spectrum caps applied in large markets such as Argentina,Brazil,Chile,Colombia,MexicoandPeru.

40

WayneA.Leighton,MeasuringtheEffectsofSpectrumAggregationLimits:ThreeCaseStudiesfromLatin America(25October2009).

41|P a g e

TheresultsofMexicanspectrumAuctions20&21heldinJuly2010servetodemonstratethe consequences of imposing spectrum caps in an era of skyrocketing demand for mobile broadband.ThespectrumcapestablishedbytheCFC(ComisinFederaldeCompetencia)for Auctions 20 & 21 meant that the maximum amount of spectrum that any operator could accumulate,includinganyspectrumheldin850MHz,1900MHzand1700/2100MHz,couldnot exceed 80 MHz. Due to the fact that three of the four operators had more than 50 MHz accumulatedinseveraloratleastoneofthenineregionswheretheauctionwasgoingtotake place,thiseffectivelymeantthatnoneofthemcouldbidforanationwideAWSlicensebecause theywouldthenexceedthe80MHzcapinoneormoreoftheregionsmonitoredbytheCFC. Theendresultisthatofthe90MHzavailableinthisbandfor3nationwidelicenses,onlyone licensewasawardedtoanewconsortium,whileonelicensewentwithoutasinglebid. Similarly, Colombias auction of spectrum in the 25002690 MHz band in July 2010 demonstratestheflawsofbluntlyappliedspectrumcaps.TheColombiangovernmentimposed a new spectrum cap of55 MHz for the established and new operators in the 850 MHz, 1900 MHzand25002690MHzbandsANDalsomandatedthatminimumbidswouldneedtobeat least30MHz.Inpractice,thismeantthatthethreeestablishedoperatorscouldnotparticipate, andthatthisspectrumwasofflimitstothoseoperators.41 Insummary,spectrumcapsinplaceinsomecountriesinLatinAmericawillinhibitanddelaythe deployment of mobile broadband services. Caps can actually limit competition by restricting output and preventing mobile operators from growing and innovating. Several more flexible regulatoryalternativeshavebeenimplementedbyothercountriestoaddresspotentiallackof competition and to achieve universal service goals without having to depend upon distortive overall spectrum aggregation limits. In general, casebycase review of a markets competitiveness,asreflectedbytheabsenceofanysingleproviderbeingabletocontrolprice orrestrictoutput,isapreferabletool.Suchanapproachwillbetterallowformobilebroadband tobedeployedtoconsumersthroughouttheAmericas.

ColombiasMinistryofICT,ithasbeenreported,recentlydecidedtoraiseitsspectrumcapto60MHz. http://www.telecomsinsight.com/file/95347/colombiacreatescapacityformobilebroadbandgrowth.html. Tellingly,thisreportalsoindicatesthatdatafromtheMinistryshowsthatmobiledatausagehasgrown exceptionallyinColombia.

41

42|P a g e

4.2.1ILLUSTRATIONS

UnitedStates In2001,theFederalCommunicationsCommission(FCC)decidedtoeliminatespectrumcapsof 55 MHz (in all geographic service areas) and replace them with a casebycase competitive reviewtoensurecompetition.Ascreeningguidelinewasadopted,atwhichlevelanoperators spectrumholdingsandsituationcouldbereviewedforpotentialanticompetitiveeffects.Since theeliminationofspectrumcaps,theU.S.hasledtheworldinmobilebroadbanddeployment. Presently, the screening guideline is defined as between 95 145 MHz, depending on the availabilityofAWS1and/or2.6GHzBRSspectruminagivenmarket. Canada In 2004, Industry Canada (IC) eliminated its spectrum caps after finding that the policy to overseespectrumconcentrationhadbecomelessrelevant. UnitedKingdom

CompetitionintheU.K.mobilemarkethasbeenenabledbytheissuanceofseparatelicenses ratherthanbytheimpositionofspectrumcaps.ThesectorregulatorOfcomhaspronouncedin favor of applying general competition policy and relaxing restrictions on spectrum use to resolve or preempt potential competition problems, accompanied by bandspecific spectrum caps that are loose and flexible. Ofcom is planning an auction of the 2.6 GHz band in 2012 accordingtothefollowingguidelines: Technologyandserviceneutrality Alooseorsafeguardbandspecificspectrumcapof80MHz(outof190MHz) Norolloutorcoverageobligations Acquiredspectrumwillbetradable

43|P a g e

4.3THEREISNOTIMETOLOSESPECTRUMALLOCATIONSCANTAKEYEARSTO EFFECTUATE

Foravarietyofreasons,theprocessofallocatingspectrumhastraditionallytakenfiveormore yearstocomplete.IntheUnitedStates,forexample,themajorbandsformobilebroadband servicehavetakenanywherefromsixtothirteenyearstocompletethereallocationprocess,as illustratedinthechartbelow.

Band Cellular (Advanced Mobile Phone System) PCS Educational Broadband Service (EBS) Broadband Radio Service (BRS) 700 MHz AWS-1 First Step 1970 1989 1996 Available for Use 1981 1995 2006 Approximate Time Lag 11 years 6 years 10 years

1996 2000

2009 2006

13 years 6 years

[Source:FCCBroadbandPlan,March2010]

Whilecircumstanceswillcertainlyvarybycountryandbyspectrumband,onecanreasonably conclude from prior history and current circumstances that the process will unfold generally along similarly timescales and certainly so if the process to identify spectrum for mobile broadbandisnotundertakenwithakeensenseofurgency.

44|P a g e

4.3.1ILLUSTRATIONS

Mexico On 2 September 2010, Mexican President Felipe Caldern issued a decree to accelerate the countrystransitionfromanalogtodigitalTV,freeingup700MHzfrequenciescurrentlyusedby broadcasterswiththeintentiontoauctionthisspectrumby2012.Underthedecree,allanalog TVtransmissionsaretomigratefullytodigitalstandardsby2015,sixyearsaheadoftheoriginal plan. Cofetel has been instructed to auction the 700 MHz band for next generation mobile broadbandservices,withpreliminaryexpectationsofUS$10billioninauctionreceipts.Mexico isthefirstcountryinLatinAmericatoannounceaDigitalDividendspectrumauctionandunder its proposed schedule will also be the first country in the region to conclude the analog transition. Moreover,inearlyJanuary2011,Cofetelannouncedthatitwasworkingtorelease300MHzof spectrumfromthe700MHz,1.7/2.1GHz,and3.43.7GHzbandswithintwentyfourmonths. Thiswouldbethemostambitiousprogramtoreleasespectrumeverconducted,andwouldif accomplishedresultinadoublingofthespectrumcurrentlydedicatedtowirelessservices.42 Brazil OnDecember2010Anatelconcludedtheauctionofaleftoverblockof20MHzofthe1.9/2.1 GHz band which was initially auctioned in December 2007. The amount received by the government for this auction was US $ 1.7 billion dollars for 13 lots of the so called HBand. Eleven(Lots14,67and913)ofthethirteenlotswerewonbyNextelBrasil,onelot(#5)by CTBC,andanotherlot(#7)byTIM.

42

http://eleconomista.com.mx/industrias/2011/01/05/avanzaplancadenastv

45|P a g e

5.CLOSINGCONSIDERATIONS

Mobile broadband in the Americas is in a delicate state. On the one hand, the growth in subscribershiphasbeenphenomenal,amobilemiracleinthewordsoftheITU.Ontheother hand, the industry lacks sufficient incremental supply of one of its essential raw materials spectrum. Our review of the literature leads us to the conclusion that mobile broadband networks will hit capacity shortages by the middle of the decade unless steps are taken to securetheadditionalspectrumneeded. Such steps need to be taken today to avoid these risks. The industry has a long history of drivinginnovationinradioaccesstechnologies,fromEDGEthroughHSPAtoLTE,allowingitto exploitspectrumassetsasintensivelyaspossible.Inparallel,theindustryhasinvestedbillions inbuildingcellsitestoenhancenetworkcoverageandcapacity.Suchstepswillcontinuetobe needed,andthereisnoindicationofdeploymentslowing. Atthesametime,incrementalspectrumallocationsformobilebroadbandarevital.Countries mustbeginnow,iftheyhavenotdonesoalready,toplanforthefutureinordertopreserve thepromiseofthemobilemiracle.Historically,spectrumallocationscantakeatleast5years, andoftenlonger,toimplement. 4G Americas offers the following guideposts to help stakeholders in the region in working togethertosecureabrightmobilebroadbandtomorrow. 1.WellConsideredSpectrumAllocationPoliciesareImperative ConfigureLicenseswithWiderBandwidths GroupLikeServicesTogether BeMindfulofGlobalStandards PursueHarmonized/ContiguousSpectrumAllocations ExhaustExclusiveUseOptionsBeforePursuingSharedUse NotAllSpectrumisFungibleAlignAllocationwithDemand 2.Marketorientedspectrumassignmentapproachesworkspectrumcaps shouldbedisfavored. 3.Thereisnotimetolosespectrumallocationscantakeyearstoeffectuate. 4GAmericas,aswellasitsmembercompanies,standpreparedtoaidstakeholdersinthe regioninsecuringthepromiseofmobilebroadband. A. B. C. D. E. F.

46|P a g e

6.APPENDICES

6.1ABBREVIATIONS 3GPPThirdGenerationPartnershipProject ANATELAgnciaNacionaldeTelecomunicaes(Braziliantelecomregulator) AUAntennaUnit Bits/s/HzBitsperSecondperHertz,ameasureofspectralefficiency bpsBitsperSecond CAGRCompoundAnnualGrowthRate CAPEXCapitalExpenditure CDMACodeDivisionMultipleAccess CFCComisinFederaldeCompetencia(competitionagencyinMexico) COFETELComisinFederaldeTelecommunicaciones(telecomregulatorinMexico) CRTCCanadianRadioandTelecommunicationsCommission dBDecibel dBmDecibelratioofwattsto1milliwatt ECEuropeanCommission EDGEEnhancedDataRatesforGSMEvolution EUEuropeanUnion FCCFederalCommunicationsCommission FDDFrequencyDivisionDuplex FEMFrontEndModule FTPFileTransferProtocol GBGigabyte GbpsGigabitsperSecond GGSNGatewayGPRSSupportNode GHzGigahertz GPRSGeneralPacketRadioService GSMGlobalSystemforMobilecommunications GSMAGSMAssociation HSPAHighSpeedPacketAccess(HSDPAwithHSUPA) HSPA+HighSpeedPacketAccessPlus(alsoknownasHSPAEvolutionorEvolvedHSPA) HzHertz ICIndustryCanada IMTInternationalMobileTelecommunications IMTAIMTAdvanced IPInternetProtocol ITUInternationalTelecommunicationsUnion KbpsKilobitsperSecond LTELongTermEvolution(evolvedairinterfacebasedonOFDMA) LTEALTEAdvanced M2MMachinetoMachine

47|P a g e

MbpsMegabitsperSecond MHzMegahertz MIMOMultipleInput/MultipleOutput msMillisecond OBIFCCsOmnibusBroadbandInitiative OFCOMU.K.communicationsregulatoryauthority OFDMOrthogonalFrequencyDivisionMultiplexing OFDMAOrthogonalFrequencyDivisionMultipleAccess(airinterface) OPEXOperatingExpenses P2PPeertoPeer RANRadioAccessNetwork Rel.XRelease99,Release4,Release5,etc.of3GPPStandards RFRadioFrequency RXReceive RNCRadioNetworkController RRURemoteRadioUnit SAESystemArchitectureEvolution,alsoknownasEPC SGSNServingGPRSSupportNode SNRSignaltoNoiseRatio TETerminalEquipment TXTransmit UEUserEquipment ULUplink UMTSUniversalMobileTelecommunicationsSystem UNCTADUnitedNationsConferenceonTradeandDevelopment WANWideAreaNetwork WWWWorldWideWeb WiMAXWorldwideInteroperabilityforMicrowaveAccess

48|P a g e

6.2SELECTEDREFERENCES Akamai,StateoftheInternetQ32010(January2011),availablewithfreeregistrationat http://www.akamai.com/stateoftheinternet/ AllotCommunications,MobileTrendsReportH22010(8February2011),availablewithfree registrationathttp://www.allot.com/index.aspx?id=4070&fileID=3710 Berry,HonigandVohra,SpectrumMarkets:Motivations,Challenges&Implications(IEEE CommunicationsMagazine,November2010). CiscoVisualNetworkingIndex:GlobalMobileDataTrafficForecastUpdate,201015(1 February2011), http://www.cisco.com/en/US/solutions/collateral/ns341/ns525/ns537/ns705/ns827/white _paper_c11520862.pdf FCC,NationalBroadbandPlan,Chapter5:Spectrum(March2010), http://www.broadband.gov/plan/5spectrum/#_edn67 FCC,OBITechnicalPaperNo.6,MobileBroadband:theBenefitsofAdditionalSpectrum(21 October2010),http://download.broadband.gov/plan/thebroadbandavailabilitygapobi technicalpaperno1.pdf ITU,EstimatedSpectrumBandwidthRequirementsfortheFutureDevelopmentofIMT2000 andIMTAdvanced,ReportITURM.2078(2006). NGMNAlliance,SpectrumRequirementsfortheNextGenerationofMobileNetworks(20 June2007), http://www.ngmn.org/uploads/media/Spectrum_Requirements_for_the_Next_Generation _of_Mobile_Networks.pdf DrMartynRoetteronbehalfoftheGSMA,SpectrumforMobileBroadbandintheAmericas: PolicyIssuesforGrowth&Competition(January2011),availableat http://www.gsmamobilebroadband.com/upload/resources/files/GSMA%20Americas%20M BB%20Spectrum%20Paper%20Jan2011.pdf. RTTEconomics,LTEUserEquipment,NetworkEfficiency&Value(September2010), availablewithfreeregistrationathttp://www.makingtelecomswork.com/resources.html RysavyResearch,MobileBroadbandCapacityConstraintsReport(February2010), http://www.rysavy.com/Articles/2010_02_Rysavy_Mobile_Broadband_Capacity_Constraint s.pdf Sandvine,Fall2010GlobalInternetPhenomenaReport(20October2010),availablewith freeregistrationat http://www.sandvine.com/news/global_broadband_trends.asp#theform UMTSForumReportNo.44,MobileTrafficForecasts20102020Report,(January2011).

49|P a g e

7.ACKNOWLEDGEMENTS

The mission of 4G Americas is to promote, facilitate and advocate for the deployment and adoptionofthe3GPPfamilyoftechnologiesthroughouttheAmericas.4GAmericas'Boardof Governor members include AlcatelLucent, Amrica Mvil, AT&T, Cable & Wireless, Commscope, Ericsson, Gemalto, HP, Huawei, Motorola, Nokia Siemens Networks, Openwave, Powerwave, Qualcomm, Research In Motion (RIM), Rogers, Shaw Communications, TMobile USAandTelefnica. 4G Americas would like to recognize the significant project leadership and important contributionsofBobCalaffofTMobileUSA,aswellasrepresentativesfromtheothermember companies on 4G Americas Board of Governors who participated in the development of this whitepaper.

50|P a g e