Voting

Edward Felten, Andrew Appel, Joseph Hall

Appropriate use of technology in the voting process is an area of active research at the Center. Center members consider both the promise of electronic voting technologies and their limitations. Past projects at the Center have revealed flaws and security vulnerabilities in deployed voting machines. To counter these threats, researchers here are exploring techniques such as auditing to help ensure trustworthy elections even when conducted with imperfect technology. Center members have provided technical guidance on this topic to elected officials, courts, and other governmental bodies.

Transparency

Edward Felten, Matt Salganik, David Robinson

The Center has several projects under way that explore Internet technologies as tools for government transparency. A recent paper offered recommendations for the federal executive branch, while other ongoing research relates to citizen participation in online deliberative processes.

Community-Generated and Community Sorted Information

The goal of this research is to develop a suggestion box for the digital age, a tool that will enable groups to collect and then collaboratively evaluate new information such as ideas or photos. This new method is based on the old-fashion suggestion box which has proven to be both practical and flexible. With a suggestion box, however, all that one has at the end is a pile of suggestions with little idea of the preferences of the group for these suggestions. Now imagine that while these suggestions are coming in, the researcher could take subsets of suggestions and distribute them to people in the community to be voted upon. If these distributed voting tasks were correctly organized and efficiently aggregated they could be used to prioritize suggestions, as well as reveal other information about the preferences of the group. This research will develop the appropriate algorithms to distribute and aggregate votes, as well as develop a set of open source software tools that will make these methods available to millions of groups all over the world such as local governments, firms, religious congregations, and NGOs.

Enabling untrusted content distribution for the Web

Michael Freedman, Ed Felten

Content distribution networks (CDNs), whether commercial CDNs or peer-to-peer (P2P) systems, provide the heavy-lifting for Web content distribution. Yet today’s browser is built around the identifying primitive of a domain name. Conflating naming, location, and authorization, browsers use domains both to specify where to retrieve content and what security policies to enforce on downloaded objects. This design conflation makes all third-party CDNs inappropriate for secure Web pages (https) and, moreover, peer-to-peer systems unsuitable for any Web content. In doing so, it greatly raises the bar for content publishers to reach large audiences, and further inhibits the use of secure transmission protocols.

This research argues for the move from a domain-based Web security model to a signature-based one. Our server and browser extensions enable the secure exchange of web content via potentially untrusted proxies, whether belonging to traditional CDNs or even P2P ones. Domains can carefully specify which third-parties are delegated content distribution responsibility. In the peer-to-peer model, users can enforce privacy preferences by carefully specifying which content they will share. The protocols are backwards compatible with today’s web standards, integrate easily with existing web servers, and are designed not to interfere with a typical browsing experience and
publishing ecosystem.

Greening IT: From the Phone to the Data Center

Mung Chiang, Michael Freedman, Margaret Martonosi, Jennifer Rexford

Information (and communications) Technology (IT) consumes 2.5% of the world’s electricity, translating to one billion tons of CO2 annually. For example, just in the United States, data centers alone consume more than 60 billion kilowatt hours a year—equivalent to that consumed by the entire transportation manufacturing sector. In this project, we investigate methods to reduce IT energy consumption from an end-to-end perspective—from the mobile phone to the data center. Our approach is interdisciplinary, drawing on our combined expertise in computer architecture, communication networks, and wide-area services, drawing on both theoretical techniques and systems prototyping. We investigate the challenges with providing workload-proportional usage; supporting workload migration both across the local and wide area; using prediction versus reaction approaches for managing, consolidating, and migration jobs; and exploring the trade-offs between energy, performance, and reliability.

Computer Networking and Public Policy

Jennifer Rexford

This work addresses a policy audience to foster innovation in computer networks, seeking to create a future Internet that is secure, reliable, efficient, and economically-viable. Areas of interest include technological surveillance and crisis-level traffic burdens on the Internet. More information here.

Digital Technologies and Social Inequality

Paul DiMaggio

Paul DiMaggio studies the relationship between the new digital technologies and social inequality, focusing on the Internet in the United States.. To what extent is there a “digital divide,” what are the mechanisms that drive it, and is it likely to be a transitional or persistent problem? How, and through what mechanisms, does social inequality influence the use that people make of new technologies once they begin to use them? What evidence do we have that technology use influences (or does not influence) people’s life chances, social resources, and participation in community and political affairs? Recent work addresses the ways in which network externalities and social homophily interact to produce long-term intergroup inequality in technology use. He also teaches an undergraduate course with computer scientist David Dobkin on public policy issues related to the Internet at the Woodrow Wilson School.

The Mobile Phone as a Social Science Research Tool

Thomas Espenshade & John Palmer

This project examines the potential of location-aware mobile telephoney for gathering social science research data.