Guest

Visual Networking Index (VNI)

Cisco Visual Networking Index: 2013–2018 Forecast Q&A

  • Viewing Options

  • PDF (264.2 KB)
  • Feedback

June 2014

Q. Why did Cisco develop the Cisco Visual Networking Index (Cisco VNI ) forecast and methodology?
A. The ramifications of prior Internet traffic growth rates prompted Cisco to provide arealistic forecast that isbased on multiple levels and sources of real data and projections. This data is of great interest to us, but wealso expect that our customers (in all segments) and the industry at large can benefit from our findings.
Q. What is visual networking?
A. Consumer and business IP networking trends are largely generated by video and by social networking, along with collaboration, such as Web 2.0 technologies. This combination is termed visual networking. A visual networking experience can range from a telepresence meeting to the delivery of video toany device a consumer chooses, such as a TV, PC tablet, or smartphone.
Q. When did Cisco begin forecasting global IP traffic?
A. The Cisco VNI began in 2006. In that year the company published its report internally, butalso shared the forecast and projections with customers and prospects. Based on the interest in and effects of the initial report, Cisco began releasing these findings publicly in 2007.
Q. What is the methodology behind the Cisco VNI Forecast?
A. The forecast relies on analyst projections for Internet users, broadband connections, video subscribers, mobileconnections, and Internet application adoption. Our trusted analyst forecasts come from Kagan, Ovum,Informa, Ookla Speedtest.net, IDC, Gartner, ABI, AMI, Strategy Analytics, Infonetics, Dell’Oro, Synergy, Screen Digest, Pyramid, comScore, Nielsen, Arbitron Mobile, Media Partners Asia, Machina, andavariety of other sources.

In addition, a number of service providers share network traffic data and trends with Cisco, and this data isused tovalidate and adjust the usage assumptions underlying the forecast model.

Q. Have there been any methodological changes since the last forecast update?
A. The following enhancements have been incorporated into the methodology since the last update:

Country-level forecast data for Sweden and Saudi Arabia has been added to the forecast. In previous forecast updates, thisdata was included in the “Rest of Western Europe” and the “Rest of Middle East andAfrica” categories.

Projections for global IP traffic (particularly video) transported over content delivery networks (CDNs) wereupdated in this iteration of the forecast.

Global analysis and comparison of metro and long-haul IP traffic were updated in this year’s report.

Revised information on global IPv6-capable devices and connections is included in this year’s study.

Specific information on global and regional Wi-Fi speed, in addition to the standard fixed broadband speedanalysis, has been updated for the current forecast period.

Q. Were there any changes to the basic assumptions or key influences on IP traffic used in the latest forecastupdate?
A. In one area, independent data sources were revised and restructured in the Cisco VNI regional distribution ofInternet users. Specifically, Turkey was moved from the Middle East and Africa Region to the Central and Eastern Europe region. This change effectively reduced the current and projected number of users in the Middle East and Africa while increasing that current and future category for Central and Eastern Europe.

Also, as we reported in the February 2013 Cisco VNI Mobile Data Traffic Forecast, the slowed growth of PCand laptop devices (particularly in Western Europe and North America) led to a reduction in the global andregional device projections. Business desktops and consumer desktops and laptops in particular are experiencing negative growth rates. Conversely, tablets continue to be among the fastest growing device types in each region, but not at a rate that directly replaces the slowed or decreased rate of PC and laptopdevices.

Broadband speeds for Australia and New Zealand have been lowered due to the updated fiber to the home/building (FTTH/B) projections, based on recent developments with those countries’ National BroadbandInitiatives.

Q. What’s the difference between IP traffic and Internet traffic?
A. The Cisco VNI IP traffic forecast encompasses both Internet and managed IP or non-Internet IP traffic. “Internet traffic” includes all IP traffic that crosses an Internet backbone, often traversing networks operated bydifferent service providers and content delivery network providers. “Managed IP traffic” is limited to IP trafficthat is managed from origin to destination by a single service provider, crossing only a single network. Managed IP traffic includes the IP transport of consumer video on demand (VoD), as well as corporate IP WAN traffic. Many forecast metrics vary significantly, depending on whether they include managed IP traffic or Internet-only traffic. For example, we project that Wi-Fi traffic will represent 49 percent of total global IP traffic (which includes managed IP) by 2018. Comparatively, we estimate that Wi-Fi traffic will represent 61 percent of total global Internet traffic (which excludes managed IP) by 2018. The lower Wi-Fi percentage for overall IPis due to the prevalence of wired devices, such as set-top boxes, that generate IP VoD traffic.
Q. Have you changed any historic traffic projections in the latest forecast update?
A. Each year, we use what we believe to be the most accurate data and methodology possible to update theforecast. This year, to improve our forecasting and data collection methodology, we made the following adjustments to 2012 and 2013 IP traffic volumes: In most regions, the actual volumes for fixed Internet trafficwere adjusted downward by less than 10 percent except in the case of Latin America, where the figureswere revised more significantly, due to new bitrate measurements that show lower bit rates than previously estimated.
Q. What is the future outlook for IP traffic growth based on the updated forecast?
A. In general, all regions experienced slight reductions compared to last year’s forecasted growth rates. While for most regions the reductions were in line with the leveling of growth rates that is typical of S-curve growth, two regions had growth rates above or significantly below what was expected during the previous forecast period (2012-2017). In North America, steady growth of IP VoD led to an overall increase of projected growth rates compared to what wasexpected in last year’s forecast. In Latin America, on the other hand, the projected growth rate has beenreduced beyond what would normally be expected from S-curve decreases, in part because consensus forecasts for device growth in the region have been reduced. Latin America now has thelowest compound annual growth rate (CAGR) for devices, from 2012 to 2017, of any region.

By 2018, we will be firmly established in the zettabyte era (Figure 1). Major findings of the Cisco VNI GlobalForecast, 2013–2018, include the following:

Globally, IP traffic will grow nearly threefold from 2013 to 2018, a CAGR of 21 percent.

Globally, IP traffic will reach 131.9 exabytes per month in 2018, the equivalent of transporting 45 million DVDs each hour.

Globally, IP traffic will reach an annual run rate of 1.6 zettabytes by the end of 2018, up from an annual runrate of 614.0 exabytes at the end of 2013.

Asia Pacific will generate the most IP traffic by 2018 (47.3 exabytes per month).

The Middle East and Africa will be the fastest-growing IP traffic region from 2013 to 2018 (fivefold growth, 38percent CAGR over the forecast period). Central and Eastern Europe will be the second fastest-growing region (threefoldgrowth, 23 percent CAGR) in this category.

By 2018, non-PC devices will account for 57 percent of total IP traffic, demonstrating the effect ofweb‑enabled TVs, tablets, and smartphones on the way consumers access and use the Internet.

Mobile data traffic will be over 12 percent of global IP traffic by 2018, growing from 3 percent of global IPtraffic in 2013.

Q. What is an exabyte? What is a zettabyte?
A. An exabyte is 1,000,000,000 gigabytes. A zettabyte is 1,000 exabytes. Figure 1 shows examples of data thatreaches the exabyte and zettabyte scale.
Figure 1. The Zettabyte Scale

Q. How accurate is your forecast?
A. The Cisco VNI forecast has been characterized as conservative by some industry analysts and academicians. However, last year’s Cisco VNI projections have aligned with several actual growth reports from independent third parties.

Telegeography reported that global average international Internet traffic grew 35 percent in 2013. TheCiscoVNI Internet traffic growth rates for the primary countries included in the Cisco VNI forecast ranged from 10to 50 percent for 2013, which is consistent with the Telegeography findings.

Q. Why isn’t the projected global IP traffic growth rate as high as it has been in previous forecasts?
A. The updated global IP growth rate is not as high as in previous forecasts, but this is not a traffic reduction—it isastandard function of sigmoid curve (or S-curve) modeling. The initial stages of growth can be exponential, butthat pace cannot continue indefinitely. However, in real terms, top-level traffic growth represents a huge amount of traffic. For example, in 2018 alone, more traffic will traverse global networks than all prior “Internet years” combined. IP traffic between the years 1984 and 2013 was 1.3 zettabytes, which is lower than the annual global IP traffic of 1.6 zettabytes of traffic in 2018.
Q. How do you define your major application categories?
A. The following major application categories and corresponding definitions are used within our Cisco VNIframework.

File sharing includes peer-to-peer (P2P) traffic from all recognized P2P systems, such as BitTorrent and eDonkey, along with other means of file sharing and one-click file hosting (for example, Rapidshare).

Internet video to PC includes online video that is downloaded or streamed for viewing on a PC screen. Itdoes not include P2P downloads or Internet delivery of video to a TV screen through aset-top box or equivalent device. Internet video viewed on PCs includes a growing volume of long-form commercial content, such as movies and TV episodes, as well as short-form content such as freeuser-generated clips.

Internet video to TV includes video delivered by the Internet to a TV screen, through an Internet-enabled set-top box or equivalent device. Examples of devices and services that deliver this type of content include web-enabled TVs and Blu-ray disc players, Roku boxes, Apple TV, and gaming consoles that allow users todownload movies and broadcast television content.

Web and data include web browsing, email, instant messaging, newsgroups, and file transfer, but they donot include P2P and commercial file transfer such as iTunes. This general category encompasses data transfer (including file transfer using HTTP and FTP) and other Internet applications.

Video on demand is managed IP transport (traffic that remains within the footprint of a single service provider) generated by traditional commercial TV services, including standard definition (SD), high definition(HD), and three-dimensional (3D) cable and Internet Protocol Television (IPTV) VoD.

Q. Is there a way to customize this forecast for a specific set of variables?
A. Yes, we have two tools—one for fixed networks (CT-SPAN: www.cisco.com/go/span) and the other for mobilenetworks (MSPAN: www.cisco.com/go/mspan)—that can generate customized forecasts based oninputs you specify. These tools use Cisco VNI assumptions for defaults and can take user inputs for auser‑customized forecast.
Q. Does this forecast include signaling traffic?
A. No, signaling traffic is not included. However, an estimate can be made using the standard rule that IPsignaling traffic is approximately 3 percent of bearer traffic.
Q. Are traffic patterns becoming more symmetric over time?
A. No, we have observed that, despite the increase in consumer uploading of user-generated content and content providers making longer-form content available online, the amount of downloading still exceeds uploading, and traffic patterns are increasingly asymmetric.
Q. Does the forecast include both uplink and downlink traffic?
A. For most services, the figures reflect both downlink and uplink traffic. To avoid double-counting, we excluded uplink P2P, VoIP, instant messaging, and video calling. In other words, we did not include uplink traffic for anyapplication where one person’s upload is another person’s download. Uploads to servers (for example, YouTube) are included, however.
Q. Cisco VNI appears to be focused on forward-looking data. Is there data showing how Internet traffic has developed historically?
A. Yes, we have historical data for global and U.S. Internet traffic. Historical Internet traffic includes consumerand business Internet traffic, but it does not include traffic on business IP WANs and private andmobile networks.
Q. Why is broadcast TV traffic so low in comparison to VoD traffic?
A. Broadcast traffic is low because it is a one-to-many service rather than a one-to-one service, like VoD. Foreach VoD request, a new stream must be served. However, when hundreds of people tune in to the sametelevision show, only one copy of this show crosses most of the network. When it is near the network edge, itissplit and sent over each access line. In this forecast, the access-line traffic for broadcast TV is notincluded.
Q. What about satellite video traffic?
A. Because satellite is similar to broadcast, in that it is a one-to-many service, the exclusion of satellite from the forecast is notexpected to make a significant difference. However, direct broadcast satellite (DBS) providers are now deploying set-top box services that simulate VoD. They send the top 25 programs to the set-top boxes overnight andsupport on-demand access to less-popular content through the subscriber’s Internet connection. This on‑demand streaming is certain to have an effect on traffic in the future, and it has been factored into the assumptions for Internet video.
Q. What about digital terrestrial video traffic?
A. Like satellite, digital terrestrial television (DTT) is a one-to-many service, so the exclusion of DTT is not expected to materially affect the accuracy of the forecast. Also like satellite providers, pay-DTT providers mayestablish a broadband connection to the home so that they can offer on-demand content, Internet content, and interactive content. This scenario has not been included in this forecast, because the penetration of DTT remains low throughout the forecast period. DTT may be included in future versions of the forecast.
Q. Can my organization or I use or publish Cisco VNI Forecast data?
A. Yes. Cisco welcomes and encourages press, analysts, service providers, and other interested industry parties, whether business, regulatory, or academic, to use or publish the data. Cisco VNI projections have been cited in equity and investment research, S-1 registration statements, initial public offering (IPO) and Security and Exchange Commission (SEC) filings, and offering memorandums. We have shared our publicly published data with government regulators, press, industry analysts, academic institutions, technical conferences, journals, andother media outlets. We do require that proper Cisco attribution be given for any and all Cisco VNI datathat ispublished or shared in private or public, print and electronic forms (for example, “Source: CiscoVisual Networking Index [or VNI] Global IP Traffic Forecast, 2013–2018”). No further signatures or consent are required to refer to our publicly available white papers, reports, or web-based tools. We are always interested in the context in which our data is used and would appreciate it if parties that use our content would share copies of their completed work containing Cisco VNI insertions. Please send these to traffic-inquiries@cisco.com.
Q. Can you share the application-level traffic data and country-level data you used to construct the regional trafficfigures?
A. We are not able to share the specific source data that serves as a primary input to our forecast methodology,but we have developed an interactive tool that generates customized Cisco VNI data basedonuser requirements. The Cisco VNI Forecast Tool is an online resource that is publicly available.Youcangenerate your own Cisco VNI Forecast charts based on segment, region, country, orotherparameters. The forecast tool is available at www.ciscovni.com/vni_forecast/index.htm. Wealsohaveanadvanced highlights tool. Thistool generates important projections from the IPtrafficandmobile datatrafficforecast for each region andcountry covered by the research (referto www.cisco.com/web/solutions/sp/vni/vni_forecast_highlights/index.html).
Q. Where can I ask questions about the Cisco VNI Global IP Traffic Forecast data?
A. Please send any questions to traffic-inquiries@cisco.com.

For More Information

For more information, refer to www.cisco.com/go/vni. Inquiries can be directed to traffic-inquiries@cisco.com. Youmay also submit questions or comments to our feedback section at www.cisco.com/go/vni.