D02 Isap2b.1.1 C 0025!!PDF e

Download as pdf or txt
Download as pdf or txt
You are on page 1of 15

International Internet

Statistics

ITU ICT Indicators Meeting

February 10, 2005

www.telegeography.com

Overview

• International Internet Bandwidth


• International Internet Traffic
• International VoIP Traffic
• Data Sources and Research Challenges

www.telegeography.com Proprietary and Confidential 2

1
International Internet Bandwidth

www.telegeography.com Proprietary and Confidential 3

What is Internet Bandwidth?


• Internet bandwidth is capacity which backbone operators
provision to carry IP traffic.

• Internet bandwidth (capacity) is not the same as Internet


traffic (capacity usage).

• Internet bandwidth data do not reflect the true origin or


destination of Internet traffic. Traffic may travel between
any two points on the network.

www.telegeography.com Proprietary and Confidential 4

2
Internet is the Largest Consumer of
International Submarine Cable Capacity

www.telegeography.com Proprietary and Confidential 5

Major International Internet Routes

Notes: Map includes international routes with at least 5 Gbps of aggregate capacity. Routes scaled to reflect Internet bandwidth connect across
international borders to each country. Data as of mid-2004.

www.telegeography.com Proprietary and Confidential 6

3
International Internet Bandwidth
Growth by Region

Notes: Data as of mid-year

www.telegeography.com Proprietary and Confidential 7

Interregional Internet Bandwidth, 2004

Notes: Data as of mid-year. Interregional bandwidth below 1,000 Mbps not depicted.

www.telegeography.com Proprietary and Confidential 8

4
Highest Capacity International
Internet Routes for Asia (Mbps)

Rank City, Country 2002 2003 2004


1. San Francisco, U.S.-Tokyo, Japan 9,552 24,046 44,159
2. Los Angeles, U.S.-Tokyo, Japan 4,605 8,082 17,875
3. San Francisco, U.S.-Seoul, Korea, Rep. 5,894 10,606 16,638
4. Seoul, Korea, Rep.-Tokyo, Japan 2,319 8,978 14,103
5. Hong Kong, Hong Kong-Tokyo, Japan 2,164 5,808 13,462
Notes: Figures represent Internet bandwidth connected across international borders to Consolidated Metropolitan Statistical Areas or
equivalents as of mid -year. Domestic routes omitted.

www.telegeography.com Proprietary and Confidential 9

Highest Capacity International Internet


Hub Cities for Asia (Mbps)

Rank City 2002 2003 2004


1. Tokyo, Japan 39,605 97,351 178,605
2. San Francisco, U.S. 34,062 62,365 102,902
3. Hong Kong, Hong Kong 22,174 47,079 93,835
4. Seoul, Korea, Rep. 16,843 43,356 71,381
5. Taipei, Taiwan 16,848 28,927 65,908
Notes: Figures represent Internet bandwidth connected to Asian locations across international borders from Consolidated Metropolitan
Statistical Areas or equivalents, including cities outside of Asia. Data as of mid -2004. Domestic routes are omitted.

www.telegeography.com Proprietary and Confidential 10

5
TeleGeography’s International Internet
Bandwidth Data Set
• Data consist of international city-to-city routes in terms of
Mbps by company.

• Data set has been collected annually since 1998.

• Currently over 12,500 individual route records totaling


over 10 Tbps of Internet bandwidth are stored in our
internal database.

• For asymmetric Internet links, the downstream


bandwidth is recorded.

www.telegeography.com Proprietary and Confidential 11

International Internet Traffic

www.telegeography.com Proprietary and Confidential 12

6
What is Internet traffic?
• To measure Internet traffic, TeleGeography uses “link
utilization” - a common metric used by network
engineers to measure the performance of their networks.

• Link utilization statistics show how much Internet


capacity is being used between two cities during a given
period of time, but does not show the true origin or
destination of the traffic.

• TeleGeography collects both the average and peak


utilization levels across international links.

www.telegeography.com Proprietary and Confidential 13

International Internet Capacity Utilization

Notes: Data reflect traffic over Internet bandwidth connected across international borders including links within the region. Data as of April
2004.
www.telegeography.com Proprietary and Confidential 14

7
Latin American International Internet
Traffic by City and by Route
Internet Average Average Peak Peak
City, Country Capacity Traffic Utilization Traffic Utilization
Buenos Aires, Argentina 12,248 3,184 26% 4,401 36%
Lima, Peru 5,644 2,285 40% 3,091 55%
Mexico City, Mexico 4,974 1,711 34% 2,313 47%
Santiago, Chile 12,704 3,656 29% 3,851 30%
S‹o Paulo, Brazil 18,433 7,098 39% 9,729 53%

Internet Average Average Peak Peak


Route Capacity Traffic Utilization Traffic Utilization
Miami, U.S.-S‹o Paulo, Brazil 11,969 4,729 40% 6,490 54%
Miami, U.S.-Santiago, Chile 7,817 3,271 42% 3,262 42%
Lima, Peru-Miami, U.S. 5,286 2,160 41% 2,916 55%
Buenos Aires, Argentina-Miami, U.S. 5,280 2,208 42% 3,009 57%

Notes: Figures represent Internet traffic and bandwidth connect across international borders as of April 2004. Domestic routes are
omitted

www.telegeography.com Proprietary and Confidential 15

TeleGeography’s International Internet


Traffic Data Set
• Average and peak levels of international Internet traffic
collected since 2003 for 45 cities, 60 routes, and 45
countries.

• Data are based on a one month sample from April of


each year.

• Data are bidirectional averages - traffic flows in both


directions across an Internet link.

www.telegeography.com Proprietary and Confidential 16

8
International VoIP Traffic

www.telegeography.com Proprietary and Confidential 17

What is international VoIP traffic?


• TeleGeography tracks international phone calls that
transit public or private IP networks at some point, but
ultimately terminate on traditional fixed or mobile
networks.

• PC-to-PC communications and intra-company private


network traffic are excluded from VoIP data set.

www.telegeography.com Proprietary and Confidential 18

9
International VoIP and Switched Traffic
Growth, 1997-2004

Notes: VoIP traffic includes all cross-border voice calls carried on IP networks but terminated on publ ic switched telephone networks; PC-
to-PC communications and private network traffic are excluded. Switched traffic includes circuit-switched voice and fax traffic carried on
traditional international facilities as well as international simple resale (ISR) facilities. Figures for 2004 are estimated.

www.telegeography.com Proprietary and Confidential 19

International Inbound VoIP Traffic Totals


by Region, 2001-2003

Notes: VoIP traffic includes all cross-border voice calls carried on IP networks but terminated on publ ic switched telephone networks; PC-
to-PC communications and private network traffic are excluded.

www.telegeography.com Proprietary and Confidential 20

10
International VoIP Termination
in Africa, 2003

Notes: Total inbound VoIP traffic to Africa was one billion min utes in 2003. VoIP traffic includes all cross-border voice calls carried on IP
networks but terminated on public switched telephone networks; P C-to-PC communications and private network traffic are excluded.

www.telegeography.com Proprietary and Confidential 21

TeleGeography’s International VoIP


Traffic Data Set
• Traffic data collected annually in terms of number of
minutes.

• Minutes tracked by termination point.

• Data has been collected annually since 1999.

www.telegeography.com Proprietary and Confidential 22

11
Data Sources and Research Challenges

www.telegeography.com Proprietary and Confidential 23

Internet Bandwidth & Traffic Data Sources


• Annual confidential survey of backbone providers.
• Public and private information sources used to
supplement bandwidth data.
• Automated network discovery tool also used for
bandwidth data. This tool collects network topology data
for a large number of geographic location, which is then
parsed for identifiable geographic and commercial
characteristics.

International VoIP Traffic Data Sources


• Annual survey of VoIP carriers.

www.telegeography.com Proprietary and Confidential 24

12
Research Challenges
• Confidentiality: Data collected from carriers is
extremely sensitive, particularly Internet traffic.
Convincing carriers to contribute data each year is
difficult.

• Number of Providers: Over 400 carriers operated


international Internet links in 2004. Impossible to receive
data from every carrier. Estimation is required to create
complete Internet bandwidth and traffic data sets.

www.telegeography.com Proprietary and Confidential 25

Research Challenges
• Internet Traffic Origin and Destination: Traffic data do
not reveal origin or destination of traffic.

• VoIP Traffic Origin: Not possible to track the origin of


VoIP traffic due to hubbed routing.

www.telegeography.com Proprietary and Confidential 26

13
Global Internet Geography

TeleGeography’s
international
Internet bandwidth
and traffic data are
updated annually
in the Global
Internet
Geography report.

www.telegeography.com Proprietary and Confidential 27

TeleGeography

TeleGeography’s
international VoIP
data are updated
annually in the
TeleGeography
report.

www.telegeography.com Proprietary and Confidential 28

14
Thank You

Alan Mauldin
Senior Research Analyst
+1 202 741 0048
[email protected]

www.telegeography.com

15

You might also like