Review - Tussle in Cyberspace: Defining Tomorrow's Internet

Review of Tussle in Cyberspace: Defining Tomorrow's Internet
by Ivan Hernández

The Internet has grown dramatically in the recent years, it has become
from a research curiosity into a mainstream commodity. The main
implication of this, the Internet as an economy engine, is that there
are important and powerful players with interests that may be adverse
to each other (users, ISPs, Governments, Intellectual property right
holders, and providers of services). The paper explores the impact of
these tussles in the designs principles for network architectures.

The paper states that the technical architecture must accommodate the
tussles of society, while continuing to achieve its traditional goals
of scalability, reliability, and evolvability. If think that this is
true, in the sense that the network must be able to provide general
services with a greater granularity level that the services can use,
such as security, accounting, and identity. Nevertheless, I found that
is not possible to always constraint networking engineering solutions
to what "society or some players" want, because the output of this
will not necessarily be the most efficient or of general use.

Solutions that are less efficient from a technical perspective may do
a better job of isolating the collateral damage of tussle. It is true
that solutions are designed to satisfy particular requirements. For
instance a particular car model satisfies a particular set of
costumers. Nevertheless computer networks does not provide their
services to a particular product, but to several products some of them
were not though in the beginning. I think that this because computer
networks design has been driven mainly to provide a set of technical
requirements. On the contrary computer networks. not have been
designed to meet particular service requirements.

The paper has interesting discussion, in particular related to
economics tussles of the Internet. In section 3.1.3, the authors talk
about a pessimistic landscape for residential customers, in five years
(i.e., 2007) the average residential costumers will only have two
options to access Internet the telephone and cable
companies. Nevertheless, the emergence of 802.16 and its
interoperability forum WiMax will enable the delivery of last mile
wireless broadband access as an alternative to cable and DSL. This
will probably keep up the competition on residential Internet access.

I do not think that network designers have to modularize the design
along tussle boundaries. They must modularize along the most efficient
boundaries given the problem to solve, not given the policies. I think
that one example of a problem solved with this principle in mind could
be BGP, routing driving by policies, and have seen that BGP is not a
protocol with good performance. I work but is not the best solution.

Finally, this is more a discussion paper, it only provides some
toughs from the authors. In most of them they provide simplistic
solutions to real problems (like how to prevent DOS
attacks). Furthermore, some of their principles look like old
principles with a little variation or in words of the authors these
are "perhaps" new principles.
Received on Mon Nov 27 2006 - 20:27:16 EST