Ericsson Mobility Report June 2013
Upcoming SlideShare
Loading in...5

Ericsson Mobility Report June 2013



The new edition of the Ericsson Mobility Report reveals that mobile-data traffic will continue to grow significantly in the coming years, a trend driven mainly by video. Overall data traffic is ...

The new edition of the Ericsson Mobility Report reveals that mobile-data traffic will continue to grow significantly in the coming years, a trend driven mainly by video. Overall data traffic is expected to grow 12-fold by the end of 2018. Increasing usage is driven by continual growth in the amount of content available as well as the improved network speeds that come with HSPA and LTE development.
For more on the latest mobility figures see:



Total Views
Views on SlideShare
Embed Views



11 Embeds 931 368 325 55 54 45 41 27 10 3 2 1



Upload Details

Uploaded via as Adobe PDF

Usage Rights

© All Rights Reserved

Report content

Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

  • Full Name Full Name Comment goes here.
    Are you sure you want to
    Your message goes here
Post Comment
Edit your comment

Ericsson Mobility Report June 2013 Ericsson Mobility Report June 2013 Document Transcript

  • June 2013ON THE PULSE OF THE NETWORKED SOCIETYEricssonMobilityReport
  • 2  ERICSSON MOBILITY REPORT  JUNE 2013Mobile Subscriptions Update 4Subscriptions Outlook 6Mobile Traffic Update 9Traffic Development 10Population Coverge 14App Coverage 16Consumer Loyalty and Network Performance 18International Data Roaming 20Beyond Bytes 22The Signature of Humanity 24Key Figures 26Methodology and Glossary 27The contents of this document are based on a number of theoretical dependenciesand assumptions and Ericsson shall not be bound by or liable for any statement,representation, undertaking or omission made in this document. FurthermoreEricsson may at any time change the contents of this document at its solediscretion and shall not be liable for the consequences of such changes.CONTENTS
  • JUNE 2013  ERICSSON MOBILITY REPORT  3EricssonMobility ReportJune 2013The number of mobile subscriptions worldwide has grown approximately8 percent year-on-year during Q1 2013. The number of mobile broadbandsubscriptions grew even faster over this period at a rate of 45 percentyear-on-year, reaching around 1.7 billion. The amount of data usage persubscription also continued to grow steadily. About 50 percent of allmobile phones sold in Q1 2013 were smartphones.These factors combined have resulted in the total amount of mobiledata traffic doubling between Q1 2012 and Q1 2013. In this issue we lookbeyond our usual update and outlook on mobile traffic with an analysisof traffic types. Traffic has been forecast by application type, providinganalysis into factors behind the growth.In the last issue we highlighted the growing importance of speed in mobilebroadband networks. In this edition we take this analysis a step further,describing a new framework to understand varying network performancewithin a given coverage area in our App Coverage article. Continuing onthe theme of network performance, we cover the relationship betweenperformance and customer loyalty and explain the effects of signalingon data traffic. We also take a look at data roaming, and identify anopportunity for operators to generate new revenue streams.Our final chapter, the Signature of Humanity, offers a visualizationof data and voice activity in three major cities around the world.Finally, you will find our usual table of key figures at the end of thisdocument. We hope you find this report engaging and valuable.ABOUT THIS REPORTEricsson has performed in-depthdata traffic measurements in mobilenetworks from the world’s largestinstalled base of live networks.These measurements have beencollected from all regions of theworld since the early days ofmobile broadband.The aim of this report is toshare analysis based on thesemeasurements, internal forecastsand other relevant studies to provideinsights into the current traffic andmarket trends.Publisher: Douglas Gilstrap,Senior Vice President andHead of Strategy, EricssonSUBSCRIPTIONS 6The number ofmobile broadbandsubscriptions hasgrown by around45 percent year-on-year, reachingaround 1.7 billion.TRAFFIC 10Video traffic in mobilenetworks is expectedto grow by around60 percent annuallythrough to 2018.POPULATION COVERAGE 14LTE will coverapproximately60 percent of theworld’s populationin 2018.
  • 4  ERICSSON MOBILITY REPORT  JUNE 2013Mobilesubscriptions updateFigure 1Figure 1: Mobile subscriptions by region, Q1 2013NET ADDITIONS Q1 2013 (MILLION)MOBILE SUBSCRIPTIONS (MILLION)5322638102848775271,275271,1473071610695133643130net adDitionsQ1 2013millionMobile subscription figures are estimates as of Q1 2013. Mobilenet additions are estimates during Q1 2013. APAC = Asia Pacific.The estimate of mobile net additions has been made based onhistoric information from external sources and regulatory andoperator reports, combined with Ericsson analysis. Historicaldata may be revised when operators report updated figures.Indian subscriptions have been adjusted to reflect active subscriptionsas defined by the telecom regulatory authority in India. This better reflectsreal subscriber growth at a time when India’s operators have beendeleting inactive subscriptions.All figures are rounded up or down to the nearest million. The sum of allrounded regional data may therefore differ slightly from the rounded total.Mobile subscriptions include all mobile technologies. M2M subscriptionsare not included.6.4mobile subscriptionsglobally in Q1 2013BillionSource: Ericsson (June 2013)
  • JUNE 2013  ERICSSON MOBILITY REPORT  5Subscriptions vs subscribersThere is a large difference between the number ofsubscriptions and subscribers. This is due to the factthat many subscribers have several subscriptions.Reasons for this could include users lowering theirtraffic cost by using optimized subscriptions fordifferent types of calls, maximizing coverage andhaving different subscriptions for mobile PCs/tabletsand for mobile phones. In addition, it takes time beforeinactive subscriptions are removed from operatordatabases. Consequently, subscription penetrationcan easily reach above 100 percent, which is the casein many countries today. It should however be notedthat in some developing regions, it is common forseveral people to share one subscription, having forexample a family or community shared phone.Figure 2: Penetration percentage, Q1 2013> Global mobile penetration was at 90 percent inQ1 2013 and mobile subscriptions now total around6.4 billion. However, the actual number of subscribersis around 4.5 billion, since many people haveseveral subscriptions> China alone accounted for around 25 percent of netadditions, adding around 30 million subscriptions> India (+10 million), Indonesia (+10 million),Brazil (+5 million) and Nigeria (+5 million)follow in terms of net additions> Global mobile subscriptions have grownaround 8 percent year-on-year and 2 percentquarter-on-quarter> Global mobile broadband subscriptions1have grownby around 45 percent year-on-year and have reachedaround 1.7 billion> There is continued strong momentum forsmartphone uptake in all regions. Around50 percent of all mobile phones sold inQ1 2013 were smartphones, compared toaround 40 percent for the full year in 2012.Of all mobile phone subscriptions 20-25 percentare associated with smartphones, leavingconsiderable room for further uptake> LTE is now growing strongly, with around20 million new subscriptions added in Q1 2013.GSM/EDGE-only subscriptions added ~30 millionand WCDMA/HSPA ~60 million1Mobile broadband is defined as CDMA2000 EV-DO, HSPA, LTE, Mobile WiMAX and TD-SCDMA.Global penetration 90%Central & Eastern Europe 132%Source: Ericsson (June 2013)Western Europe 128%Latin America 114%Middle East 109%Africa 71%APAC excluding China & India 100%China 84%India 56%North America 103%
  • 6  ERICSSON MOBILITY REPORT  JUNE 2013Mobile PCs, tabletsand mobile routersFixed broadbandMobile subscriptionsSubscriptions/lines(million)Mobile broadband01,0002,0003,0004,0005,0006,0007,0008,0009,0002018201720162015201420132012201120102009SubscriptionsoutlookKey changes in this forecast:> Stronger uptake of smartphone1and LTEsubscriptions than in previous forecast> Notable increase in the number of smartphonesubscriptions in the APAC and MEA regions> LTE subscriptions forecast has increasedfrom 1.6 to 2 billion in 2018By Q1 2013, total mobile subscriptions exceeded6.4 billion. By the end of 2018, they are expectedto reach 9.1 billion.Global mobile broadband subscriptions reachedaround 1.7 billion in Q1 2013, and are predictedto reach 7 billion in 2018. The majority of mobilebroadband devices are, and will continue tobe, smartphones. Mobile broadband will gain alarger share of total broadband subscriptions inmany markets, complementing xDSL in certainsegments and replacing it in others. Mobilebroadband also includes some feature phones.Mobile devicesMobile subscriptions are increasing for PCs,mobile routers and tablets that use larger screensizes. They are expected to grow from 300 millionin 2012 to around 850 million in 2018, exceedingthe number of fixed broadband subscriptions.Total smartphone subscriptions reached 1.2 billionat the end of 2012 and are expected to grow to4.5 billion in 2018. Today the majority of mobilesubscriptions are for basic phones.Smartphone penetration will increase rapidly, whileit is estimated that subscriptions for basic phoneswill remain high, slowly declining from around 5billion today, to around 4 billion in 2018. This isbecause a large part of the growth in subscribernumbers will come from the lower-endphone segment.Figure 4: Smartphone, PC, mobile routers and tabletsubscriptions with cellular connection, 2009-201801,0002,0003,0004,0005,0006,0002018201720162015201420132012201120102009Mobile PCs, tablets and mobile routersSmartphonesSubscriptions(million)Source: Ericsson (June 2013)Mobile broadband: Includes CDMA2000 EV-DO, HSPA, LTE, MobileWiMAX and TD-SCDMA. For example, mobile PCs, tablets, routers andmobile phones, including both smartphones and feature phones. The vastmajority are mobile phones.Mobile PC: Includes laptop or desktop PC devices with built-in cellularmodem or external USB dongle.Mobile router: A device with a cellular network connection to the internet,and Wi-Fi or Ethernet connection to one or several clients (such as PCsand tablets).4.5smartphonesubscriptionsbythe end of 2018BillionFigure 3: Fixed and mobile subscriptions2, 2009-2018Source: Ericsson (June 2013)1Roughly half of the increase in the smartphone forecast vs.the previous edition is due to an increase in the underlyingdevice forecast. The rest is reflecting a modification to thedefinition of smartphone subscriptions.2Subscriptions do not include M2M subscriptions.
  • JUNE 2013  ERICSSON MOBILITY REPORT  701,0002,0003,0004,0005,0006,0007,0008,0009,00010,0002018201720162015201420132012201120102009Mobilesubscriptions(million)LANAAPACMEACEEWE01,0002,0003,0004,0005,0006,0007,0008,0009,00010,0002018201720162015201420132012201120102009Mobilesubscriptions(million)LTEWCDMA/HSPAGSM/EDGE-onlyTD-SCDMACDMAOtherFigure 5: Mobile subscriptions by technology, 2009-2018Figure 6: Mobile subscriptions by region, 2009-2018Source: Ericsson (June 2013)Source: Ericsson (June 2013)Subscriptions are defined by the most advanced technology thatthe mobile phone and network are capable of. Lack of 3G/4Gcoverage may connect users to 2G networks (GSM, CDMA1X).Regional differences will be large. In2018 almost all handsets in WesternEurope and North America will besmartphones, compared to 40-50percent of handset subscriptions inthe Middle East and Africa andAsia Pacific regions.Mobile technologyFigure 5 illustrates reportedmobile subscriptions categorizedby technology. Subscriptions aredefined by the most advancedtechnology that the mobile phoneand network are capable of.LTE is currently being deployed andbuilt-out in all regions and will reacharound 2 billion subscriptions in 2018.These subscriptions will represent thehigh-end share of the total subscriberbase by 2018. Rapid migration tomore advanced technologies indeveloped countries means globalGSM/EDGE-only subscriptionnumbers will decline after 2012-2013.Globally, GSM/EDGE will continue tolead in terms of subscription numbersuntil the latter years of the forecastperiod. This is because new, lessaffluent users entering networks ingrowing markets will be likely to usethe cheapest mobile phones andsubscriptions available. In addition,it takes time for the installed base ofphones to be upgraded.Regional growthFigure 6 illustrates regional mobile subscriptions up until the end of 2018 and shows continued growth.In APAC this process is driven by new subscribers, whereas in North America, it is based on increasingthe number of subscriptions per individual – for example, adding a tablet.The number of fixed broadband users is at least three times thenumber of fixed broadband connections, due to multiple usage inhouseholds, enterprises and public access spots. This is the oppositeof the mobile phone situation, where subscription numbers exceeduser numbers. In the latter years of the forecasting period, it is likelythat the usage trend for mobile PCs will be similar to fixed broadbandusage today, with several users per subscription. This is especiallythe case in developing markets where mobile access will be themain source of internet connection.>9mobilesubscriptionsby the end of 2018Billion
  • 8  ERICSSON MOBILITY REPORT  JUNE 2013LTEWCDMA/HSPAGSM/EDGE-onlyTD-SCDMACDMAOther020%40%60%80%100%APACMEACEEWELANASource:Ericsson (June 2013)Regional technology maturityDifferent regional maturity levelsare reflected in their radio technologymix. Less mature regions aredominated by 2G technologies,like GSM/EDGE, while more matureregions like Western Europe aredominated by HSPA. LTE is growingvery strongly, particularly in NorthAmerica. In all regions, 2G networks(GSM/EDGE, CDMA 1X) remainas fallback networks for 3G and4G subscriptions when coverageis missing.North America is characterized byearly growth in LTE. This technologywill represent the majority ofsubscriptions in the region 2016,growing to around 70 percent in2018. GSM/EDGE-only subscriptionswill progressively decline. This fastgrowth in LTE subscriptions is drivenby strong competition and consumerdemand, following CDMA operators’early decisions to migrate to LTE.Latin America has a large GSM/EDGE subscriber base. The stronggrowth in subscriptions in thisregion will be driven by economicdevelopment and consumer demand.In 2018, WCDMA/HSPA will be thedominant technology, howeverGSM/EDGE-only subscriptionswill still have a significant presence.As a mature market, WesternEurope will show little growth insubscriptions. What growth there iswill come from an increasing numberof connected devices. HSPA is thedominant technology in the region.By 2018, LTE will penetrate around35 percent of the subscriptions basein Western Europe. Data serviceswere rolled out early in this region,initially accessed via dongle or PC.The drive for LTE has not yet beenas strong in Europe, partly becausethere are many well-developed 3Gnetworks in the region.Figure 7: Mobile subscriptions by technology and region, 2012Figure 8: Mobile subscriptions by technology and region, 2018LTEWCDMA/HSPAGSM/EDGE-onlyTD-SCDMACDMAOther020%40%60%80%100%APACMEACEEWELANASource:Ericsson (June 2013)The Asia Pacific market continues to see a massive increase insubscriptions. Markets like Japan and South Korea will take up LTEsubscriptions early compared to less developed countries. China willadd substantial LTE numbers towards the end of our forecast period.CDMA will continue to grow gradually in absolute numbers, especiallyin China and Indonesia, but the CDMA share of subscriptions will decline.TD-SCDMA subscriptions will start to decline at the end of theforecast period.Central and Eastern Europe shows a strong increase in HSPAsubscriptions. LTE will initially grow in the most developed partsof the region, and will be present in most countries by 2018.The Middle East and Africa region was dominated by GSM/EDGE in2012. By 2018 it will have the largest share of GSM/EDGE, driven bydemand for low cost phones. The region is diverse, so there will belarge differences between developed and less developed areas.Subscriptions are defined by the most advanced technology that themobile phone and network are capable of. Lack of 3G/4G coveragemay connect users to 2G networks (GSM, CDMA1X).
  • JUNE 2013  ERICSSON MOBILITY REPORT  9Global traffic in mobile networksFigure 9 shows total global monthly data and voicetraffic. It depicts a stable trend of data traffic growthwith some seasonal variations. It shows that mobiledata subscriptions will grow strongly, driving thegrowth in data traffic along with a continuous increasein the average data volumes per subscription.Highlights include:> Data traffic doubled between Q1 2012 and Q1 2013> Data traffic growth between Q4 2012 and Q1 2013was 19 percent> Voice traffic growth between Q1 2012 and Q1 2013was 4 percentIt should be noted that there are large differences intraffic levels between markets, regions and operators.These measurements have been performed byEricsson over several years using a large base ofcommercial networks that together cover all regionsof the world. They form a representative base forcalculating world total traffic in mobile networks1.Figure 9: Global total data traffic in mobile networks, 2007-201304008001,2001,6002,000Q1Q4Q3Q2Q1Q4Q3Q2Q1Q4Q3Q2Q1Q4Q3Q2Q1Q4Q3Q2Q1Q4Q3Q2Q1DataVoiceTotal(uplink+downlink)monthlytraffic(PetaBytes)20132007 2008 2009 2010 2011 2012Source: Ericsson (June 2013)1Traffic does not include DVB-H, Wi-Fi, or Mobile WiMax. Voice does not include VoIP. M2M traffic is not included.Mobiletraffic update2xmobile Datatrafficdoubled betweenQ1 2012 and Q1 2013
  • 10  ERICSSON MOBILITY REPORT  JUNE 2013TrafficdevelopmentHighlights of forecast:> Smartphone traffic will have increased by ~10 percentin 2018 compared to the previous forecast> Video will account for around half of global mobiledata traffic by 2018Traffic outlookDuring 2013, overall mobile data traffic is expectedto continue the trend of doubling each year.Mobile PCs dominate traffic in most regions,except in North America. However, smartphonetraffic is growing faster due to the high growth insubscriptions. In the latter years of the forecastperiod, data traffic will be split fairly equally betweenmobile phones on the one hand, and tablets, mobilerouters and mobile PCs on the other.Mobile data traffic will grow considerably fasterthan fixed data traffic over the forecast period.However, in absolute volume, fixed data trafficwill remain dominant over the same period.Accessing the internet through dedicated appssuch as social networks and picture messagingwill drive mobile traffic development. Mobile datatraffic is expected to grow with a CAGR of around50 percent (2012-2018), driven mainly by video.This will result in growth of around 12 times bythe end of 2018.Traffic per subscriber is partly affected by thescreen size of the user’s individual device.Resolution is also a factor, with recent smartphonesclosing in on PC-level quality. On average, a mobilePC generates approximately five times more trafficthan a smartphone. By the end of 2012, an averagemobile PC generated approximately 2.5 GB permonth, versus 450 MB per month produced byMonthlyExabytes(1018)03691215201820172016201520142013201220112010Data: mobile PCs, tablets and mobile routersData: mobile phonesVoiceFigure 10: Global mobile traffic: voice and data, 2010-2018Source: Ericsson (June 2013)0306090120150201820172016201520142013201220112010MonthlyExabytes(1018)GB/month/subscription02468101220182012Mobile PCsSmartphonesFigure 11: Global fixed traffic, 2010-2018Figure 12: Smartphone and mobile PC traffic per monthand active subscriptions, 2012 and 2018Source: Ericsson (June 2013)Source: Ericsson (June 2013)Mobile data trafficMobile data traffic is a share of total fixedand mobile traffic. It represents 4 percentin 2012 and 9 percent in 2018.12xMobile data trafficwill grow by ~12 timesbetween 2012 and 2018
  • JUNE 2013  ERICSSON MOBILITY REPORT  11Trafficmanagement impactNote that a large part of data traffic(fixed and mobile) is generated bya limited number of users in eachdevice category. The forecast issensitive to operator data volumecaps and tariff changes.NAWECEMALAAPACNAWECEMALAAPAC020%40%60%80%100%APACCEMAWELANAData: mobile phonesData: mobile PCs,tablets and mobile routersVoiceFigure 13: Mobile traffic by region and type, 2012Figure 14: Mobile traffic share by region, 2012Figure 15: Mobile traffic share by region, 2018Traffic refers to aggregated traffic in mobile access networks.DVB-H and Wi-Fi traffic have not been included. M2M trafficis not included.Source: Ericsson (June 2013)Source: Ericsson (June 2013)Source: Ericsson (June 2013)the average smartphone. By the end of 2018,it is estimated that on average a mobile PCwill generate around 11 GB per month and asmartphone around 2 GB. Note that there arelarge differences in user patterns betweendifferent networks, markets and user types.Regional mobile traffic variationsThe diverse maturity levels between regionscan be seen by looking at each region infigure 13 and comparing the traffic generatedfrom different device types.As illustrated in figure 14, Asia Pacific hada high share of the total traffic during 2012.North America and Western Europe have asignificantly larger share of total traffic volumethan their subscription numbers alone wouldimply due to a high proportion of mobile PCs,smartphones and tablets in 3G/4G networks.North America, which is dominated bysmartphone traffic and already has a highpenetration of smartphones, will have a lessershare of global traffic in 2018 than in 2012.This is because the smartphone share of totalphone subscriptions – expected to alreadybe at 80 percent in 2015 – will be saturatedbefore the other regions.Data traffic doubled between Q1 2012 andQ1 2013 and will continue to grow at a high rate,whereas voice traffic will maintain moderatesingle digit growth per annum. In other words,by 2018, voice traffic volumes in all regions willbe very small compared to data traffic.
  • 12  ERICSSON MOBILITY REPORT  JUNE 2013Mobile data traffic outlook by application typeThe growth in the number of mobile broadbandsubscriptions is a strong driver for mobile data trafficgrowth. More users are adding an ever-broadeningrange of devices to their subscriptions, such assmartphones, tablets, mobile PCs/routers, eBookreaders and cameras.The fastest growing segment in mobile data trafficis video. Increasing usage is driven by continualgrowth in the amount of available content as well asthe better network speeds that come with HSPA andLTE development. Larger device screens and betterresolutions will also drive video traffic as they willenable high definition and eventually even ultra highdefinition video.Video streaming services in some markets have showna very strong uptake: people use services such asNetflix, HBO and Vimeo on all types of devices. Asvideo conferencing evolves beyond fixed facilities inmeeting rooms to being used on mobile devices, itwill also drive video traffic growth in mobile networks.Today, video makes up the largest segment of datatraffic in mobile networks and is expected to grow byaround 60 percent annually up until the end of 2018, bywhich point it is forecast to account for around half oftotal global traffic1.Music streaming is gaining popularity and audio isexpected to grow with a annual growth rate of around50 percent. There is a high degree of uncertainty inthe forecast for audio traffic at this stage, as it is verydependent on how music streaming services developover the coming years.Web browsing and social networking will eachconstitute around 10 percent of the total datatraffic volume in 2018.The arrival of new types of devices or content canrapidly change traffic patterns by adding types notcurrently considered significant. Furthermore, therewill be broad variations between networks withdifferent device profiles – for example, some willbe PC dominated while others will mainly facilitatesmartphone users. Traffic will also vary betweenmarkets due to differences in content availabilityand content rights issues.1Video is also likely to form a major part of file sharing traffic and a sizeable part of encrypted traffic, in addition to the identified applicationtype ‘video’.0246810121416201820172016201520142013201220112010MonthlyExabytes(1018)OtherEncryptedSoftware download& updateSocial networkingFile sharingWeb browsingAudioVideoFigure 16: Mobile data traffic by application type, 2010-2018Source: Ericsson (June 2013)Video will account foraround half of globalmobile data traffic by 2018
  • JUNE 2013  ERICSSON MOBILITY REPORT  13User behavior and driversWhen a user changes devices from a feature phone toa smartphone, voice and text messaging continue tobe used. Users tend to develop their application usagebehavior over time as they discover new apps andservices that capture their interest. Recommendationsfrom friends, family, media and app stores for new andtrendy apps play an important role in this development.New communication behaviors drive data traffic inmobile networks. A good example of this is sharingvideo via MMS or instant messaging.Over time, users tend to use more advanced servicesthat put greater demands on device capabilities. Today,smartphone users who subscribe to both music andvideo streaming services already consume more than2 GB of data traffic per month on average. That’s overfour times the consumption of an average smartphoneuser. In many markets, legal streaming services forboth music and video are gaining popularity. Givensufficient content, ease of use and the right pricelevel, these services exhibit strong adoption rates.The outlook for every category of mobile data showssignificant growth through to 2018. The highest growthis expected from video traffic, and it is estimated toconstitute around half of all mobile data traffic by theend of the forecast period. Video traffic, including theportion of encrypted and file sharing traffic which itforms, is likely to represent the majority of all mobiledata traffic by 2018.Figure 17: Mobile data traffic share by application type, 2012OtherEncryptedSoftware download& updateSocial networkingFile sharingWeb browsingAudioVideo21%5%9%10%14%2%31%8%Figure 18: Mobile data traffic share by application type, 2018OtherEncryptedSoftware download& updateSocial networkingFile sharingWeb browsingAudioVideo16%6%7%9%10%2%46%4%Source: Ericsson (June 2013)Source: Ericsson (June 2013)New communication behaviors drivedata traffic in mobile networks
  • 14  ERICSSON MOBILITY REPORT  JUNE 2013%populationcoverageWorldpopulationdistributionWCDMA/HSPAGSM/EDGE LTE020406080100201820122018201220182012RuralSub-urbanUrbanMetro>85%>90%~55%>85%~10%~60%POPULATIONCOVERAGEPopulation coverage of the world’s mobilenetworks is constantly increasing as morebase stations are deployed. This articledeals with trends and outlook concerningpopulation coverage – the area in whicha user has sufficient signal to connectto a mobile network.GSM/EDGE technology has by far the widestreach and today covers more than 85 percentof the world’s population. If Japan and Korea– two countries that have not deployed GSM/EDGE – are excluded, then the GSM/EDGEworld population coverage would be around90 percent. The areas that remain to becovered by GSM/EDGE in countries thatuse the technology are sparsely populated.By the end of 2012, WCDMA/HSPA covered55 percent of the world’s population. Furtherbuild-out of WCDMA/HSPA will be driven by anumber of factors: increased user demand forinternet access, the increasing affordability ofsmartphones, and regulatory requirements toconnect the unconnected. By the end of 2018,over 85 percent of the world’s populationFigure 19: Population coverage by technology, 2012-20181 The figures refer to population coverage of each technology.The ability to utilize the technology is subject to other factors as well, such as accessto devices and subscriptions.Source: Ericsson (2013)Latin AmericaN population coverageLatin America lacks a well built-out fixedtelecommunications infrastructure. Because of this,the majority of people in the region will first becomeregular internet users over mobile networks.Last year’s GSM population coverage in LatinAmerica and its subsequent forecast for 2018closely mirrors the overall global situation.The WCDMA/HSPA population coverage is somewhatlower in Latin America than the global average. It isestimated that by 2018, 80 percent of thepopulation in this region will be covered byWCDMA/HSPA networks.Brazil, Mexico and Colombia are early adoptersof LTE compared to the rest of Latin America. It isforecast that this technology´s population coveragein Latin America will increase from approximately 5percent to over 50 percent by 2018. During the sameperiod the number of LTE subscriptions is expectedto grow from about 200,000 to 95 million.will have the opportunity to access the internet usingWCDMA/HSPA networks1.Despite being in the early days of roll-out, by the endof 2012 LTE was estimated to cover 10 percent ofthe world’s population. Looking ahead 6 years, it ispredicted that LTE will cover approximately 60 percentof the world’s population.60%LTE will cover around60 percent ofthe world’spopulation by 2018
  • JUNE 2013  ERICSSON MOBILITY REPORT  15020406080100HSPA 42HSPA 21HSPA 7.2HSPA%100%~85%>60%>25%WCDMA/HSPA networksThere are over 500 WCDMA/HSPA networks thatcurrently provide coverage to approximately 55 percentof the world’s population. All of these networks supportvarious speeds. All WCDMA networks deployedworldwide have been upgraded with HSPA. Around85 percent of HSPA networks have been upgraded toa peak downlink speed of 7.2 Mbps or above and over60 percent have been upgraded to 21 Mbps or higher.Over 25 percent of HSPA networks now have speedsof up to 42 Mbps in whole or parts of the networkfollowing a wave of upgrades. We are already seeingevolutionary steps towards speeds of over 100 Mbps.A further development that needs to be consideredwhen analyzing WCDMA/HSPA network coverageis the increasing deployment of WCDMA/HSPA 900MHz networks. 900 MHz networks can complement2,100 MHz deployments to improve coverage andquality of service and enhance the user experience.In particular, WCDMA/HSPA 900 MHz provides betterindoor coverage than is possible using 2,100 MHz. Thefirst WCDMA/HSPA 900 MHz network was deployedin November 2007. Today, there is a well-establishedecosystem for terminals and there are 64 commercialWCDMA/HSPA 900 MHz networks in 42 countries2.LTE network rolloutDespite being in the early days of rollout, LTE networkscan already provide downlink peak rates of around100 Mbps, with current standardization allowing foreven higher speeds. Today, peak speeds experiencedby users are often limited by device capabilities. Theevolution of LTE, also referred to as LTE-Advanced,enables peak data rates exceeding 1 Gbps.LTE has become the fastest developing system inthe history of mobile communication due to the fastspeeds and high quality user experience that it offers.It has now been launched on all continents (exceptAntarctica), by 156 operators active in 67 countries2.Figure 20: Percentage of WCDMA networks upgradedto HSPA and to 7.2, 21 and 42 MbpsSource: Ericsson and GSA (April 2013)Rapid deployment has been facilitated by the abilityto re-farm spectrum and the utilization of multi-standard radio solutions. Re-farming enables anoperator to quickly make new spectrum availablefor LTE, without the need to acquire new bandwidth.Multi-standard radio solutions enable rapid deploymentof mixed technologies, enabling GSM/WCDMA/HSPAand LTE to be provided within the same radiobase station.The user experience can be further enhanced byaggregating multiple carriers across different bands.This enables the delivery of higher data rates acrosscells. Carrier aggregation is especially important forthe user experience in performance-challengedareas, such as the cell edge.2 Global Supplier Association (GSA) estimates, Q1 2013.25%Over 25 percent of HSPAnetworks are capableof speeds up to 42 Mbps
  • 16  ERICSSON MOBILITY REPORT  JUNE 2013The explosive growth of smartphone and app usagehas put a new focus on cellular network performance.App coverage is a term used to describe thisperformance in a way that is relevant to the user.The concepts in this article will deal with varyingperformance levels within the area ofpopulation coverage.Does a network enable a good experience when aparticular app is being used? This question can onlybe answered in the context of the app and the demandit places on the network in terms of connection time(or time to view), throughput and latency. From a userperspective, app coverage can be defined as the areain which the network performance is sufficient to runa specific app at an acceptable quality level.Figure 21 illustrates network performance decliningwithin a cell in relation to distance from the radio basestation. In reality, mapping network performance ismore complex, due to motion as well as geographicaland man-made obstructions to the radio signalsbetween the base station and a mobile device.The level of network performance any given userexperiences is also affected by the number of otherusers active in the cell and the demands their appsplace on the network at any one time.For social media or web browsing, the user experienceis affected by how fast requested pages are renderedonscreen. For every second that passes between auser clicking a link and a page loading, the probabilityincreases that the user will give up waiting.Apps which involve streaming media such as videohave even more stringent performance demands thanweb browsing. These apps use a range of mechanismsto smooth over variable network performance, suchas adaptive codecs, local caching and buffering. If thethroughput is not sufficient, time to view grows and theapp may periodically freeze until it is no longer possibleto stream, and then stop functioning altogether.While downlink throughput is the limiting factor formany apps, real-time apps such as video conferencingput additional requirements on uplink and latency inorder to provide sufficient quality for a gooduser experience.Network performance requirementsRequirements on network performance are both variedand complex. Given today’s devices and apps, we cangeneralize and gain an idea of what level of downlinkthroughput is necessary for a good experience.App Coverage10Mbps0.1Mbps1MbpsFigure 21: A conceptual view of app coverageSource: Ericsson 2013App coverage –the proportionof a network´s coverage thathas sufficient performanceto run a particular app at anacceptable quality level
  • JUNE 2013  ERICSSON MOBILITY REPORT  17WesternEuropeCentral& EasternEuropeMiddleEast AfricaAsiaPacificNorthAmericaLatinAmerica0.1 Mbps 94% 82% 94% 90% 85% 95% 84%1 Mbps 78% 55% 77% 58% 54% 78% 48%10 Mbps 17% 5% 27% 7% 8% 31% 5%Figure 22: Proportion of Speedtest measurements with a given downlink throughput or higher, March 2013Source: Based on Ericsson’s analysis of results provided by Ookla (2013)For simple web browsing, email and instantmessaging, 100 Kbps would be sufficient.1 Mbps would be sufficient for audio and videostreaming as well as social media with multimediacontent embedded.A downlink throughput of 10 Mbps delivers a verygood user experience, enabling higher qualityvideo streaming and real-time video conferencing.These downlink speeds are indicative, andapply to typical devices and apps in use today.Requirements such as time-to-view, uplinkthroughput and latency should also be takeninto consideration.The radio technologies used to build today’smobile broadband networks are capable ofdelivering very high performance in terms ofthroughput and latency. The challenge is to buildout enough bandwidth at site-to-site distancesto enable a high probability of getting sufficientperformance throughout each cell.Devices and appsThe device may be a factor limiting performance.A smartphone that supports up to 7.2 Mbps willnot enable the user to experience the higherspeeds available on today’s HSPA networks.A device without LTE support cannot connectto a 4G network, regardless of the technology’scoverage build-out. On the other hand, devicesare replaced at a very fast rate.Technological development is continuous. 3 yearsago, high-end smartphones with 4-inch screensand 1 GHz processors supported WCDMA/HSPA 7.2. Today, handset vendors are launchinghigh-end smartphones with 1.5 GHz quad-coreprocessors supporting downlink speeds of upto 100 Mbps on LTE networks.1 The Speedtest app from is available for devices running either IOS or Android, allowing users to measure uplink and downlinkthroughput as well as latency. Each time the app is run, the results are stored in a database by Ookla, enabling statistical analysis.The databaseis growing rapidly and approaching 1 billion measurement records.Some of these handsets feature screens offive inches or more with double the pixel densitycompared to the previous generation. There is acorresponding increase in performance for thehigh-volume, low-end and mid-range smartphonesegments. Developers are designing apps to utilizethese new capabilities. As these apps emerge, it willfuel user demand for better network performance.A new approachIn the smartphone era, there is a seemingly infinitevariety of apps. Each has its own demands for networkperformance in terms of throughput and latency.The traditional way of viewing coverage – an areathat meets a set of fixed requirements – is no longerenough. A new approach that takes a diverse set ofperformance demands into consideration is needed.Using data from Speedtest1for the month of March2013, figure 22 details the percentage of measurementswith a given downlink throughput or greater. Thestatistics come from all radio access technologiesexcept WiFi. It shows that most smartphone usersexperience sufficient network performance for voice,basic web browsing, messaging and email. However,there is a need for significantly more app coverageto run some of the more demanding apps, especiallythose that involve streaming or real-time video.As app coverage is based on user experience, allthings that affect network performance ultimatelymust be taken into consideration. This includesbackhaul capacity from the radio base stations aswell as the packet core and content delivery networks.App coverage requires a true end-to-end approachto designing, building and running mobile networks.
  • 18  ERICSSON MOBILITY REPORT  JUNE 2013Consumer Loyaltyand NetworkPerformanceNo two operators are the same,and improving loyalty meansevaluating and addressing importantparameters such as customer care,pricing and network performance.An analysis of the latter reveals thaton average, around 30 percent ofusers experience problems everyday when using their smartphonesto browse the web or use apps.As many as 60 percent experienceproblems weekly. An analysis ofthe most frequent smartphoneissues indicates that users perceivenetworks as sometimes not meetingexpected performance standards.Content needs to appear quickly onthe screen in order to ensure a gooduser experience. The most frequentlyreported complaints are about slowbrowsing, downloading and uploadingof content, followed by lack of accessto the mobile network. It has been found that 15-20 percentof users state that they experience problems very often anda further 45-50 percent experience problems occasionally.Their service expectations are simply higher than theservice level that they receive.Overall satisfaction with network performance is rankedlower than that for other important factors such as (in orderof consumer satisfaction) initial purchase experience, billingand payment, account management, devices offered andcustomer support. Value for money and tariff plans offeredare also among the important aspects where satisfaction isranked lower.Requirements on network performance are both variedand complex. Although such requirements are usuallydiscussed in technical terms, ultimately the focus is onpeople accessing networks as part of everyday activities.Whereas the App Coverage chapter on pages 16 and 17looks at how the consumer experience of coverage mayvary when using smartphone apps, the focus here is onhow that experience impacts loyalty towardsmobile operators.Lower brand perception for mobile operatorsAn analysis of brand perception reveals thatconsumers rank operators lower than other mobileecosystem players. Figure 23 shows this analysisfor one developed market only – however the globalpattern is similar.Great network – great loyaltyConsumer loyalty suffers when service factors suchas reasonable waiting times and sufficient coverageare unsatisfactory. Net Promoter Score (NPS1) is acommon metric used by various industries to measurebrand loyalty. By breaking this measurement down intodifferent factors, we can understand which areas aremore important.Figure 24 shows an analysis of NPS results. It revealsthat network performance is the principal driver ofsubscriber loyalty to mobile operators, followed byvalue for money. These two parameters are correlated,implying that improving network quality will alsoHigherLowerConsumerbrandperceptionOperatorOperator AOperator BOperator COperator DOperator EOperator FOperatingsystemSystem ASystem BSystem CSystem DHandsetmanufacturersManufacturer AManufacturer BManufacturer CManufacturer DManufacturer EManufacturer FContentprovidersProvider AProvider BProvider CProvider DProvider EProvider FProvider GProvider HProvider HFigure 23: How consumers rank brand perception in a developed marketSource: Ericsson ConsumerLab (2013)Base: Mobile phone owners in one developed market
  • JUNE 2013  ERICSSON MOBILITY REPORT  19improve perceived value for money at a given pricelevel. Spending less on network performance andpassing on savings to the consumer may alsoimprove perceived value for money, but loweringprices provides little sustainable differentiation.Figure 24 shows that addressing networkperformance has twice the impact on customerloyalty compared to measures such as improvingcustomer support. It is four times as effective asintroducing loyalty rewards.Promoters recommend their operatorFigure 25 shows that network performance isthe key factor differentiating promoters fromdetractors. The analysis also shows that up to67 percent of promoters are very satisfied withnetwork performance, compared to only 5 percentof detractors.Although operators often promote attractivehandsets to customers, figure 23 indicates that thismostly generates positive perception for the devicebrand. Instead, improving network performancemay offer a sustainable advantage, as well asincreasing an operator’s share of promoters.Figure 25: Discriminant analysis, showing differences between promoters and detractorsOverall performance for data services is perceived tobe lower than expected by consumers and so thereare opportunities for improvement, such as increasingapp coverage. Addressing network performance mayin itself be a way to differentiate offerings and also aneffective way to improve overall customer loyalty.%051015202530Handset/devicesofferedAccountmanagementCustomersupportLoyaltyrewardsBillingandpaymentOngoingcommunicationInitialpurchaseTariffplansofferedValueformoneyNetworkperformanceCustomer serviceOfferMarketingNetworkSource: EricssonConsumerLab (2013)Base: 9,040 smartphoneusers in BR, CN, SK,JP, US, UK, SE, RU, IDDiscriminant analysis, showing which satisfaction factors contribute the most to the difference in network operator advocacy.Methodology: Online sample of 9,040 smartphone users in Brazil, China, South Korea, Japan, USA, UK, Sweden, Russia, Indonesia aged 18-69.The sample is representative of 350 million smartphone users globally.Loyalty rewardsInitial purchaseHandset/devices offeredBilling and paymentAccount managementCustomer supportTariff plans offeredOngoing communicationValue for moneyNetwork perfomance 20%16%11%10%9%8%7%7%7%5%Figure 24: Drivers of loyalty to operator brand (NPS)Shapley regression analysis, showing the relative impact betweeneach driver and loyalty to operator brand (NPS).1 NPS is a quantification of the likelihood that a customer would recommend a company’s product or service. Its customers are scored on a scale of0-10 and then divided into three categories:• 9-10 – promoters: loyal enthusiasts who keep buying from the operator and referring others, fueling growth• 7-8 – passives: satisfied but unenthusiastic customers who are vulnerable to competitive offers• 0-6 – detractors: unhappy customers who can damage your brand and impede growth through negative word of mouthTo calculate an operator’s NPS, the percentage of detractors is subtracted from the percentage of promoters.Customer serviceOfferMarketingNetworkSource: EricssonConsumerLab (2013)Base: 9,040 smartphoneusers in BR, CN, SK,JP, US, UK, SE, RU, ID
  • 20  ERICSSON MOBILITY REPORT  JUNE 2013Internationaldata roamingIn this section we analyze the share ofsubscribers who enable data roaming whentraveling abroad and explore how devicepenetration differs between local subscribersand data roamers. We refer to data roamingas using mobile network data services inanother country.According to BEREC1, international dataroaming grew by 630 percent in the EuropeanUnion (EU) during 2007-2012. During thisperiod the average price per MB decreased.This, together with increased smartphonepenetration, explains the growth in dataroaming traffic.Ericsson global measurements show that theshare of data traffic volume from roamingsubscribers is lower than 0.5 percent of totalmobile data traffic. There is great variationbetween the average per subscription trafficvolumes for roaming users and local users. Insome countries it can be as low as 10 percent,while in others – such as tourist destinations –it can be on par with local usage.Figure 26 shows the share of roamingsmartphone subscribers that have dataroaming enabled and consume data. Thisshare is by far the highest on BlackBerrys(65-80 percent) followed by iPhones(20-40 percent) and Android (10-20 percent).Deeper analysis reveals that low-end Androidmodels have only ~10 percent or even lessdata roaming share, while high-end Androidsmartphones have a share similar to that ofiPhones at ~30 percent. The low share of dataroaming in figure 26 for Android smartphonesmay be due to the high penetration of low-endmodels in many networks. Most subscriberswith such a device do not have a data roamingservice or have disabled it. Some do not evenhave a data subscription available on theirhome network.For networks within the EU, the share of roaming subscriberswith an iPhone or Android smartphone that have data roamingenabled is significantly higher for roaming subscribers fromother EU countries, than from non-EU countries. This is mostprobably due to lower data roaming tariffs within the EU. Also,the majority of roaming subscribers on networks in the EU arefrom other countries within the region.User behavior changes when roaming. While video dominateslocal subscribers’ traffic it becomes negligible for those whoroam. In contrast, the relative share of maps and email trafficis significantly higher for roaming subscribers.Regional roaming patternsFew national regulators share data roaming information but theones that do – Sweden and Switzerland for example – show thatroaming patterns can differ a lot. In Sweden, 3 percent of dataroaming traffic comes from non-EU countries compared to23 percent in Switzerland. Share of local and roaming datatraffic can also vary a lot within countries. For example, oneEuropean operator which had a 28 percent market share of totalmobile traffic, had 66 percent of data roaming traffic and hencea larger share of the associated data roaming revenues.1 Body of European Regulators for Electronic Communication.010%20%30%40%50%60%70%80%Android smartphoneiPhoneBlackBerry smartphoneFigure 26: Share of roaming smartphone subscribers using data roaming servicesNote: In this analysis, only 3G traffic is considered. 3G capable deviceswhere subscribers turned off 3G capabilities and only generated 2G trafficare excluded.Network with highest share of data roamersMedian amongst measured networksNetwork with lowest share of data roamersSource: Ericsson (June 2013)
  • JUNE 2013  ERICSSON MOBILITY REPORT  21010%20%30%40%50%60%70%80%M2MFeaturephoneSymbianphoneBlackBerryphoneWindowsphoneAndoidphoneiPhoneeBookreaderTabletMobilePCRouterFigure 27: Device penetration for data roaming subscribersFigure 28: Device penetration for local mobile broadband subscribersBusiness users dominate roamingFigures 27 and 28 show that devicepenetration for data roaming subscribersdiffers from that of their local counterparts.Some devices are over-represented amongdata roaming subscribers – examples ofthis are BlackBerry smartphones, iPhonesand M2M devices. In contrast, mobile PCs,Android smartphones, Symbian smartphonesand feature phones are under-represented.Business subscribers are less sensitive tohigh roaming tariffs. This helps to explainthe high BlackBerry penetration among dataroamers in these measurements.The dominance of business users is alsohighlighted in the relatively small seasonaldifferences in device penetration for dataroaming subscribers, with no significantpeak in vacation periods.Device distribution for data roamingsubscribers is more uniform than for localsubscribers because the former come fromall over the world, and therefore regionaldifferences are leveled out. The largeruniformity can be seen in figure 27 as thevariance is less for most devices comparedto the variance in figure 28. The dominanceof business users among data roamers isanother factor that makes device penetrationmore uniform.In contrast to smartphones, the penetrationof M2M devices among data roamingsubscribers shows a broad spread acrossnetworks. This is probably due to differentlevels of penetration among M2M devicesin different regions and variations in transittraffic intensities. An example of this is whenheavy cargo transit traffic utilizes vehicletracking and fleet management M2M devices.Islands and central areas (in contrast toborders) of large countries tend to generateless data from roaming M2M devices.010%20%30%40%50%60%70%80%M2MFeaturephoneSymbianphoneBlackBerryphoneWindowsphoneAndoidphoneiPhoneeBookreaderTabletMobilePCRouterData roamingis dominated bybusiness usersNetwork withlargest penetrationof data roamersMedian amongstmeasured networksNetwork withlowest penetrationof data roamersSource: Ericsson (June 2013)Source: Ericsson (June 2013)The results in this article suggest that data roaming is dominatedby business subscribers. Data roaming penetration for consumersis lower, especially for low-end smartphone devices. This showsthat there is an opportunity for operators to generate new revenuestreams by making it attractive for subscribers that travel abroadto use their data roaming services.The measurements in this section were made in a selectednumber of commercial WCDMA/HSPA networks in Asia,Africa, the Middle East, Europe and the Americas.
  • 22  ERICSSON MOBILITY REPORT  JUNE 2013020%40%60%80%100%SmartphoneTabletMobile PCOtherEmailEncryptedSoftware download& updateSocial networkingWeb browsingLocation servicesAudioReal-timecommunicationsVideoFile sharingBeyond bytesFigure 29: Application mobile data traffic volumes by device typeFigure 29 includes 4G, 3G and 2G mobile data traffic and does not take intoaccount Wi-Fi offload traffic. Smartphones include Android and iPhone modelsonly. ‘Other’ includes applications that were not possible to identify or that don’tqualify as one of the listed applications.Video isthe biggestcontributor totraffic volumesToday, the Byte is the most widely usedmetric to characterize mobile data traffic.However, other important aspects oftraffic need to be considered whenbuilding and operating networks. Forexample, the signaling load of trafficfrom different applications impactsdimensioning of mobile data networks,while time spent using differentapplications is an important metricwhen analyzing consumer behavior.In this section, we analyze both theshare of mobile data traffic volume bydevice type, as well as traffic for differentapplications by Bytes per subscriber. Wealso introduce Minutes of Network Use(MoNU1), a metric estimating device timespent using different online applications.MoNU also correlates with signaling load –in other words, higher MoNU usually meansa greater signaling load contribution. It isimportant to note that signaling load alsovaries significantly between networkvendors and node configuration.Mobile data traffic volumes by device typeFigure 29 shows how the most widely usedonline applications contribute to overallmobile data traffic volumes, and how thesecontributions vary by the type of connecteddevice. The chart shows average valuesfrom measurements in a selected numberof commercial HSPA and LTE networks inAsia, Europe and the Americas. Actual valuesacross individual networks can vary a lot.Regardless of device type, video is the largestcontributor to traffic volumes (30-40 percent).For smartphones, social networking is alreadythe second largest traffic volume contributorwith an almost 15 percent average share.Traffic drawn from mobile PCs is notable for having significantlyhigher file sharing activity than other devices, while online audioand email are important contributors to data traffic on tablets andsmartphone devices. The file sharing part under smartphones andtablets comes from tethering traffic.Social networking a high MoNU contributorFigure 30 shows traffic for different applications by Bytes persubscriber. Figure 31 shows the same information for MoNU. Interms of Bytes, video is the largest contributor in most networksboth for mobile PC and smartphone (iPhone or Android) users.However, in terms of MoNU, video is lower and social networkingand web browsing are the largest contributors. For smartphones,presence – a background process that ensures the smartphoneSource: Ericsson(June 2013)1 Minutes of Network Use is measured by counting the number of minutes when a given terminal generates any data traffic using a givenapplication.The MoNU metric also counts usage for background applications that generate background traffic without any user interactionsuch as presence. In contrast, usage for offline applications such as offline gaming is not included. Hence MoNU is not equal to screen timebut can be a good approximation of actual screen time for most interactive applications.
  • JUNE 2013  ERICSSON MOBILITY REPORT  23MB/month/subscription05001,0001,5002,0002,5003,000SoftwaredownloadIMandpresenceEmailSocialnetworkingWebbrowsingVideoFilesharingFigure 30: Average monthly traffic volumes2 Minutes spent gaming on smartphones can be even greater than on social networking. However, many of these games are offline and do notgenerate any data traffic during most of the usage – high bytes,low MoNUMinutes/month/subscription02,0001,5001,0005002,5003,0003,5004,000SoftwaredownloadIMandpresenceEmailSocialnetworkingWebbrowsingVideoFilesharingMobile PC SmartphoneFigure 31: Average Minutes of Network Use (MoNU)largest contributor in terms of Bytes on all terminals. However,in terms of MoNU, and not counting non-interactive backgroundapplications such as presence, social networking leads onlineusage. This is especially the case with smartphones2. Non-interactive background applications on smartphones, such aspresence, are key drivers of signaling load in mobile networks.Signaling load impactDimensioning of mobile datanetworks is impacted by signalingload. Higher MoNU usually means agreater signaling load contribution.Mobile PC SmartphoneNetwork withlargest averagetraffic volumeMedian amongst measured networksNetwork withlowest averagetraffic volumeSource: Ericsson (June 2013)Source: Ericsson (June 2013)is reachable for push notifications (new email,Facebook message, chat message, etc.)– is the biggest MoNU contributor.It is interesting to compare mobile PC vs.smartphone usage for the three main interactiveonline application categories: video, webbrowsing and social networking. For video andweb browsing, mobile PC usage is higher thansmartphone usage both in terms of Bytes andMoNU. Larger screens boost usage for theseapplications. In contrast, differences in socialnetworking are much smaller and in terms ofMoNU, smartphone usage can even exceedmobile PC usage. The reason for this is that forsocial networking users, always having the devicewith them is a more important factor than screensize. It is important to note that Facebook is alsowidely used for non-social networking purposes,such as authentication for various online gamesand enriching phone contacts with social content.A significant part of Facebook MoNU on somesmartphones can be attributed to such integrationwith other apps.The spread between different terminal typesin terms of total MoNU (for all applications) issignificantly smaller than in terms of Bytes. Routerdevices and advanced smartphones with regularbackground activity have the highest average persubscription MoNU. This figure is lower for mobilePCs, which are usually only turned on when used.Due to continuous presence traffic onsmartphones, daily profiles for MoNU and radiosignaling are significantly flatter than daily profilesfor traffic volume in Bytes. This is because mostpeople do not turn off their smartphones at night.Based on our measurements, application usage andnetwork impact varies a lot on different terminalsdepending on the traffic metric used. Video is the
  • 24  ERICSSON MOBILITY REPORT  JUNE 2013THE SIGNATUREOF HUMANITY01/08/201301/07/201301/06/201301/05/201301/04/201301/03/201301/02/201301/01/201312/31/201212/30/201212/29/201212/28/201212/27/201212/26/201212/25/201212/24/201212/23/201212/22/201212/21/201212/20/201212/19/201212/18/201212/17/201212/16/201212/15/201212/14/201212/13/201212/12/201212/11/201212/10/2012ChristmasDayBoxingDayWeekendWeekendWeekendWeekendNewYear’sEveFigure 32: A comparison of voice and data traffic in London,NewYork and Hong Kong1As digital technologies become more widespread,data from communication networks allows us to betterunderstand human behavior. The study of this dataprovides new perspectives.Both similarities and differences are revealed whencomparing network usage around the globe. We seethat Londoners vary their usage a lot depending onthe day of the week, and also take a vacation frommobile networks during holidays. On the other hand,the pace in New York and Hong Kong does not changeas much. This suggests that New York has morein common with Hong Kong than London when itcomes to network usage, despite cultural differences.Nevertheless, with lower usage during the night thanthe day, people’s daily routines follow a familiar patternin all three places, and thus share a common pulse.The data used in this article was collected betweenDecember 10, 2012 and January 8, 2013, from mobilenetworks in London, New York and Hong Kong. Itshows where and when people accessed the networksto make voice calls, to use data and to send textmessages (SMS).New Yorkers stay onlineFigure 32 compares network usage in the threecities. Voice and data traffic volumes are plottedrelative to the maximum in each geographic area.The figure indicates that there is an activity dropduring weekends and holidays. Another interestingpoint revealed in the data but not illustrated in thegraphs is that the activity drop is more evident in thecity centers than in the suburbs. Most dramatic is thereduced activity in London’s financial center, whichcan likely be explained by the large number of peoplecommuting out of the area after work hours. What isvisible in the graph is that in general, New Yorkersmaintain a level of mobile activity at home, both forvoice and data, similar to that at work. In Hong Kong,while voice traffic declines in the evenings, the patternof data activity continues into the night. This is evidentin the data side of the graph in figure 32, where themeasurements are far from reaching zero in betweenthe daily peaks. People in Hong Kong use mobilebroadband to access the web and watch video clipsand movies during the night. This illustrates activitypatterns that can be matched with geographicallocation and associated culture.1Note that December 26 is a holiday in the UK and Hong Kong, but not in the US.Greater LondonNewYorkHong Kong101VoiceData101101Source: Ericsson and MIT Senseable City Lab (2013)
  • JUNE 2013  ERICSSON MOBILITY REPORT  25Wake up to dataFigure 33 compares voice, data and SMSactivity for Greater London. It plots thetime when each activity typically starts,defined as the point when the trafficreaches half its daily peak. It illustratesthat in Greater London data is frequentlyused earlier during the day than voiceand SMS. Voice activity usually beginsaround office hours at 8am, while dataactivity begins earlier at 7am, whenpeople commute to work, checkingFacebook, emails or weather reports.SMS, however, tends to be used laterthan both voice and data, except forChristmas morning and New Year’s Day.On the 31st of December, New Yorkersand Londoners wish friends and familya happy new year with an SMS. Thiscauses a predictable SMS storm. InHong Kong, the Chinese New Year hasan even larger impact on SMS traffic(however this takes place over severaldays in February, outside the boundariesof this study).The signature of humanity is a glimpseof collective behavior patterns that reveala common pulse. The pulse itself is thepoint, derived not from the content ofthe communications, but simply fromthe volume and timing of network usage.A PARTNERSHIPON THE PULSEThis article was written byEricsson in cooperation withMIT Senseable City Lab.ChristmasWeekendWeekendWeekendWeekendNewYear’sEve6:007:008:009:0010:0011:0012:0013:0014:0001/08/201301/07/201301/06/201301/05/201301/04/201301/03/201301/02/201301/01/201312/31/201212/30/201212/29/201212/28/201212/27/201212/26/201212/25/201212/24/201212/23/201212/22/201212/21/201212/20/201212/19/201212/18/201212/17/201212/16/201212/15/201212/14/201212/13/201212/12/201212/11/201212/10/2012HourofdayFigure 33: A comparison of voice, data and SMS activity for Greater LondonDataVoiceSMSSource: Ericsson and MIT Senseable City Lab (2013)
  • 26  ERICSSON MOBILITY REPORT  JUNE 2013Key figuresMobile subscription essentials 2011 20122018forecastCAGR2012-2018 UnitWorldwide mobile subscriptions** 5,800 6,300 9,100 6% millions– Smartphone subscriptions 850 1,200 4,500 25% millions– Mobile PC, tablet and mobilerouter subscriptions250 300 850 20% millions– Mobile broadband subscriptions 1,100 1,600 7,000 30% millions– Mobile subscriptions, GSM/EDGE-only 4,200 4,300 2,100 -10% millions– Mobile subscriptions, WCDMA/HSPA 950 1,200 4,300 25% millions– Mobile subscriptions, LTE 9 65 2,000 75% millionsMobile traffic essentials* 2011 20122018forecastCAGR2012-2018 Unit– Monthly data traffic per smartphone* 300 450 1,900 25% MB/month– Monthly data traffic per mobile PC* 1,900 2,500 11,000 30% MB/month– Monthly data traffic per tablet* 450 600 3,100 30% MB/monthTotal monthly mobile data traffic 600 1,200 14,000 50% PetaByte/monthTraffic growthMultiplier2012-2018CAGR2012-2018All mobile data 12 50%– Smartphones 13 55%– Mobile PC 6 35%– Tablets 40 85%PetaByte = 10^15*active devices**Using active VLR subscriptions for India*Monthly data traffic volumes by year end.
  • JUNE 2013  ERICSSON MOBILITY REPORT  27MethodologyglossaryForecast methodologyEricsson performs forecasts on a regular basis tosupport internal decisions and planning as well asmarket communication. The subscription and trafficforecast baseline in this report is based on historicaldata from various sources, validated with Ericssoninternal data, including extensive measurements incustomer networks. Future development is estimatedbased on macroeconomic trends, user trends(researched by Ericsson ConsumerLab), marketmaturity, technology development expectationsand documents such as industry analyst reports,on a national or regional level, together withinternal assumptions and analysis. Updates to thesubscription and traffic forecasts are announcedon an annual basis.Traffic measurementsNew devices and applications affect mobile networks.Having deep and up-to-date knowledge of the trafficcharacteristics of different devices and applications isimportant when designing, testing and managing mobilenetworks. Ericsson regularly performs traffic measurementsin over 100 live networks in all major regions of the world.Detailed measurements are made in a selected numberof commercial WCDMA/HSPA and LTE networks withthe purpose of discovering different traffic patterns. Allsubscriber data is made anonymous before it reachesEricsson’s analysts.2G: 2nd generation mobile networks3G: 3rd generation mobile networksAPAC: Asia PacificCAGR: Compound Annual Growth RateCDMA: Code Division Multiple AccessCEE: Central and Eastern EuropeCEMA: Central and Eastern Europe, Middle East and AfricaDSL: Digital Subscriber LineEDGE: Enhanced Data Rates for Global EvolutionExaByte: 1018BytesFDD: Frequency-Division DuplexGB: GigaByteGSA: Global Supplier AssociationGSM: Global System for Mobile CommunicationsHSPA: High Speed Packet AccessHTTP: Hypertext Transfer ProtocolIMEI-TAC: International Mobile Equipment Identity –Type Approval CodeLA: Latin AmericaLTE: Long-Term EvolutionM2M: Machine-to-MachineMB: MegaByteMBB: Mobile Broadband. Defined as CDMA2000 EV-DO, HSPA,LTE, Mobile WiMAX and TD-SCDMAMbps: Megabits per secondMEA: Middle East and AfricaMMS: Multimedia Messaging ServiceMobile PC: Defined as laptop or desktop PC devices with built-incellular modem or external USB dongleMobile router: A device with a cellular network connection to theinternet and Wi-Fi or ethernet connection to one or several clients(such as PCs or tablets)NA: North AmericaOS: Operating SystemP2P: Peer-to-PeerPetaByte: 10^15 BytesTDD: Time-Division DuplexTD-SCDMA: Time Division-Synchronous Code Division Multiple AccessVoIP: Voice over IP (Internet Protocol)WCDMA: Wideband Code Division Multiple AccessWE: Western EuropexDSL: Various technologies for DSL
  • EAB-13:028853 Uen, Revision A© Ericsson AB 2013Ericsson is a world-leading provider of communicationstechnology and services. We are enabling the NetworkedSociety with efficient real-time solutions that allow us all tostudy, work and live our lives more freely, in sustainable societiesaround the world.Our offering comprises services, software and infrastructurewithin Information and Communications Technology fortelecom operators and other industries. Today 40 percent ofthe world’s mobile traffic goes through Ericsson networks andwe support customers’ networks servicing more than 2.5 billionsubscriptions.We are more than 110,000 people working with customersin more than 180 countries. Founded in 1876, Ericsson isheadquartered in Stockholm, Sweden. In 2012 the company’snet sales were SEK 227.8 billion (USD 33.8 billion). Ericsson islisted on NASDAQ OMX, Stockholm and NASDAQ, New Yorkstock exchanges.The content of this document is subject to revision withoutnotice due to continued progress in methodology, design andmanufacturing. Ericsson shall have no liability for any error ordamage of any kind resulting from the use of this document.EricssonSE-126 25 Stockholm, SwedenTelephone +46 10 719 00 00Fax +46 8 18 40