CAREER: Predicting the Development of Decentralized Communication Systems

University Of Illinois At Urbana-Champaign, Urbana IL

Investigators

Abstract

This project will combine social scientific research on infrastructure development and the diffusion of innovations to generate a predictive spatial model of unlicensed spectrum use based on human factors. Advances in wireless technology and the present market success of "Wi-Fi" (IEEE 802.11a/b/g) wireless data networking have combined to convince many that unlicensed spectrum - spectrum that does not require permission to transmit - is now a viable way to manage a much larger amount of our wireless communication. By definition, however, use of unlicensed spectrum is unplanned and decentralized. The literature has characterized this as "chaotic" and unpredictable, and has warned that a zero price for spectrum will quickly lead to congestion failure. This project hypothesizes that unlicensed spectrum is very predictable and likely viable. Approaches to spectrum planning so far have considered few human factors. For instance, population density is often considered, but only in a simplistic manner. Hypotheses assert that spectrum in cities is more likely to be "full," and that in rural areas it is more likely to be "empty." However, many high-density areas are populated by economically disadvantaged people who may be unlikely to use a technology like Wi-Fi. This research will strive for accurate spatial predictions that include the consideration of a complex of several human factors that are not simply a matter of socioeconomic status. This project will gather and map wireless Internet use in selected census tracts in the Midwestern US and combine these data with demographics from the US Census. While the diffusion of innovations literature is well-developed, spatial analysis of diffusion is rarer and has the potential to transform our thinking about diffusion of new technologies. In addition, the project will develop the methodology of wireless mapping for social science, a promising new approach. Most significantly, the research will reconsider theories of communication infrastructure development in the light of important new technology and institutional arrangements, such as unlicensed spectrum, while building the scientific understanding of wireless data networking specifically. Whether unlicensed is a viable means for managing the spectrum is currently a pressing policy question this research can address. In addition, it should be possible to predict "holes" in use that provide the basis for innovative policymaking in the area of universal access. For example, the findings might suggest geographic spectrum easements in economically depressed areas (rather than the current thinking that privileges empty areas). The empirical data would also answer practical questions on such issues as how prevalent low-powered unlicensed transmitters are and how close they are to each other. The integrated teaching component of this project promotes a new kind of student with multidisciplinary expertise across engineering, communication, and policy. A new project-based course will be taught to teach students to collect the wireless data. As a service project successive years of the course will develop a demonstration Internet "mapserver" that allows people to visualize, test and refine predictions of spectrum use by entering their addresses, thereby promoting public understanding of science and technology.