This report forecasts how much traffic will be carried for the period to the end of this decade, on the Internet backbone and on the total national North American. Given this statement of purpose for the report, it is interesting to note that there really is no Internet backbone or North American network, in the sense of an individual entity. In the past, one could, indeed, speak of and measure the backbone traffic on the North American network. The old AT&T reported traffic loads between its Class 1 offices and the relative traffic carried on its high-usage subnet. These were real numbers that could be very accurately measured, monitored, and used as the basis of forecasts for growth.

This has all changed with the transformation to a data-based network with multiple owners. Now one is hard pressed to identify exactly what the Internet backbone is, much less to accurately measure the traffic carried on it.

Given this state of confusion and lack of physical reality of an Internet backbone, one may question the value of a report on forecast traffic. However, while the physical reality is no longer identifiable and measurable, the amount and growth of traffic going across our (somewhat imaginary) Internet backbone is still very real and of great importance. It is the growth of this traffic that

• Determines the need for equipment additions by the multitudes of carriers contributing to the various segments of this network of networks
  
• Supports facility additions (fiber and fiber routes)
  
• Requires additions to cable company networks (for high-speed) data
  
• Defines the need for higher-speed accesses and all the equipment associated with providing that extra speed
  
• Supports the ever-increasing need for safety of data, continuity of service, and privacy of data
  
• Suggests the growing value of advertising and similar activities on the Internet
  

For all of these reasons, knowledge of the future of traffic on the segments of the network and of total traffic is indispensable to all those involved in making plans for the network and all of its subparts.

We recently published a report, “How Much Bandwidth Is Enough in the Access Network” (available from IGI), addressing the question of how much bandwidth service providers needed to plan for in the last-mile plant. That report evaluated end-user tendencies to use bandwidth for video (primarily) and data access (voice being so small that it is only considered peripherally). This report is a companion report that considers the big picture of how much traffic is carried on the backbone network. The “How Much is Enough …” report is concerned with micro issues; this report is concerned with macro issues.

This report begins with a discussion of our basic approach to forecasting Internet traffic. It continues with our:

• New forecasts for high-speed access growth
  
• New forecasts for the high-speed (xDSL, cable modem, and RF) lines and traffic from high-speed access lines
  
• New forecasts for the various segments of the usage on access lines (email, searches, file sharing, instant messaging, and miscellaneous, as well as subdivisions of some of these)
  
• New FTTP lines and traffic forecast
  
• New dial-up data lines forecast and associated traffic forecast
  
• New international traffic forecasts
  
• Forecasts for various special segments — VoIP, other data networks, private lines networks
  
• Long-distance voice

Finally, the report brings all of the parts together for total network forecasts. As the report proceeds through this list of traffic sources, it provides sketches to illustrate the location of the traffic segment in the network. Major conclusions from the report are presented in a separate section.

The Appendixes provide a discussion of IPTV and a forecast for IPTV traffic. Also, extensive material is in the Appendixes, to help the reader with the various traffic concepts in this report.