Content, Centralization, and Advertising
Projects such as Programmable Open Mobile Internet contemplate the possibilities of pervasive access to the Internet via versatile mobile devices. Depending on the architecture, this may require new decisions about where to store content, and how to pay for the resources consumed when accessing content. These effects will ultimately shape and be shaped by the behavior of users.
The aim of this project is to explore the interactions between the location of content, the means by which resources are funded, and the patterns that describe how users interact with content. For instance, the choice between centralization (e.g. cloud computing) and decentralization (local storage) involves trade-offs between efficiencies of centralization (scale, redundancy) and its downsides (latency, congestion). Further, since different parties bear the costs, those costs may be recovered in different ways, leading to different kinds of contracts between the parties. Users may consume content differently depending on which contracts are present -- if advertising is used, we would expect different consumption than if users paid for a license for each item consumed.
We aim to study this problem formally, and answer questions such as the following:
- How does the location of the content impact the contracts used?
- How does the choice of contract impact the behavior of users?
- Which structures do we expect to emerge naturally?
- How does this change with the relative cost of bandwidth to storage?
Publications & Talks
D. DiPalantino and R. Johari, ``Traffic engineering versus content distribution: A game theoretic perspective,'' Proc. IEEE INFOCOM, 2009.
D. DiPalantino, R. Johari, and G. Weintraub, ``Competition and Contracting in Service Industries,'' Submitted, 2010.