Tussle in Cyberspace: Defining Tomorrow's Internet

Paper Review: Tussle in Cyberspace: Defining Tomorrow?s Internet

The paper examines the Internet process known as ?the tussle?, which
explores the dynamic between different parties with competing
interests. The paper focuses on Internet design principles, by
proposing new ones as well as revisiting the originals to accommodate
the growing instances of tussles. They argue that the problem of
tussles in the Internet has a major sociological component, rather
than being strictly an engineering problem.

The authors describe the competing parties in the Internet including
end-users, ISPs and content providers as well as some of the tussles
between them. They set out two specific design principles which aim
to permit variation in outcomes. The first requires modularity along
the tussle boundaries so that a specific tussle doesn?t affect
unrelated issues. The second states that the system be designed for
choice allowing the different parties to express their preferences.

One example in the Internet today violating the modularity design
principle is the DNS system with its dual use for names (trademarks
and machine names). SMTP is given as an example of a protocol which
is designed for choice.

The paper then raises some issues that result from the design principles:

- to enable choice, open interfaces are required
- tussles tend to occur across interfaces
- whether or not the consequences of choices are visible has an impact
- tussles come in different flavours and some have simpler solutions
than others
- tussles are dynamic and tend to evolve over time

Next, the paper examines tussles from the aspect of economics, trust
and openness. From an economic standpoint, the freedom for consumers
to choose from producers promotes competition which in turn
disciplines the market, drives innovation and removes the need for
regulation. Also, due to the growth of the Internet, there is a lack
of trust between users of the system. There exists a tussle between
regular users and the so-called bad guys (malicious users) of the
Internet leading to innovations such as the firewall.

The paper then continues by addressing some of the old Internet design
principles. The end-to-end argument led to a simple, transparent
network. Developments such as firewalls, ISP desire for control,
caches, etc., however, have stripped away at the transparency of the
Internet. They argue that the goal still remains to support new
applications as well existing maturing applications.

One complaint I have about this paper is that the introduction section
was insufficient, especially since the paper does not follow the
standard networking research paper model. Although the introduction
generates interest, it was unclear what the paper set out to
accomplish, what type of research they conducted and what were the
basic results of their work. This paper is very interesting since it
takes a completely novel approach to Internet design principles.
Rather than focusing on technology and applications, it explores the
motivation behind the parties involved. The Internet is driven by the
various players in the system and therefore it is essential to not
ignore any of the key parties and to understand the needs of each.
This paper interestingly examines the human-aspect of today?s Internet
and its potential impact on the future direction. Although the
organization of the paper was somewhat confusing, it did raise several
interesting points and thus makes a contribution to networking
research. This paper makes me realize that a project can make a
contribution without measurement, testing or simulation. Network
engineers should turn to this paper to have a sense of the human
element behind the systems they attempt to support and to better
understand what drives the interactions between these systems.

-- Nadeem Abji
Received on Tue Nov 28 2006 - 00:19:56 EST