The Internet is a network of networks that interconnects computers around
the world, supporting both business and residential users. In 1994, a
multimedia Internet application known as the World Wide Web became
popular. The higher bandwidth needs of this application have highlighted
the limited Internet access speeds available to residential users. Even at 28.8
Kilobits per second (Kbps)—the fastest residential access commonly
available at the time of this writing—the transfer of graphical images can be
This report examines two enhancements to existing residential
communications infrastructure: Integrated Services Digital ...view middle of the document...
This section does not discuss other aspects of Internet technology in
detail; that is well done elsewhere. Rather, it focuses on the services
that need to be provided for home computer users to connect to the
ISDN and upgraded cable networks will each provide different functionality
(e.g. type and speed of access) and cost profiles for Internet connections. It
might seem simple enough to figure out which option can provide the needed
level of service for the least cost, and declare that option "better." A key
problem with this approach is that it is difficult to define exactly the needed
level of service for an Internet connection. The requirements depend on
the applications being run over the connection, but these applications are
constantly changing. As a result, so are the costs of meeting the applications'
Until about twenty years ago, human conversation was by far the dominant
application running on the telephone network. The network was
consequently optimized to provide the type and quality of service needed for
conversation. Telephone traffic engineers measured aggregate statistical
conversational patterns and sized telephone networks accordingly.
Telephony's well-defined and stable service requirements are reflected in the
"3-3-3" rule of thumb relied on by traffic engineers: the average voice call
lasts three minutes, the user makes an average of three call attempts during
the peak busy hour, and the call travels over a bidirectional 3 KHz channel.
In contrast, data communications are far more difficult to characterize. Data
transmissions are generated by computer applications. Not only do existing
applications change frequently (e.g. because of software upgrades), but
entirely new categories—such as Web browsers—come into being quickly,
adding different levels and patterns of load to existing networks.
Researchers can barely measure these patterns as quickly as they are
generated, let alone plan future network capacity based on them.
The one generalization that does emerge from studies of both local and wide-
area data traffic over the years is that computer traffic is bursty. It does not
flow in constant streams; rather, "the level of traffic varies widely over
almost any measurement time scale" (Fowler and Leland, 1991). Dynamic
bandwidth allocations are therefore preferred for data traffic, since static
allocations waste unused resources and limit the flexibility to absorb bursts
This requirement addresses traffic patterns, but it says nothing about the
absolute level of load. How can we evaluate a system when we never know
how much capacity is enough? In the personal computing industry, this
problem is solved by defining "enough" to be "however much I can afford
today," and relying on continuous price-performance improvements in digital
technology to increase that level in the near future. Since both of...