CAREER: Predicting the Development of Decentralized Communication Systems

该项目将结合基础设施发展和创新传播的社会科学研究，生成基于人为因素的无牌频谱使用的预测空间模型。无线技术的进步和目前市场上“Wi-Fi”（IEEE 802.11a/b/g）无线数据网络的成功结合在一起，使许多人相信，未经许可的频谱——不需要传输许可的频谱——现在是管理大量无线通信的可行方法。然而，根据定义，使用未经许可的频谱是计划外的和分散的。文献将其描述为“混乱”和不可预测，并警告说，频谱的零价格将很快导致拥塞失败。这个项目假设，未经许可的频谱是非常可预测和可能可行的。迄今为止，频谱规划的方法很少考虑人为因素。例如，经常考虑人口密度，但只是以一种简单的方式。假设认为，城市的频谱更有可能是“满的”，而农村地区的频谱更有可能是“空的”。然而，许多高密度地区居住着经济上处于不利地位的人，他们可能不太可能使用Wi-Fi等技术。这项研究将力求准确的空间预测，包括考虑几个复杂的人类因素，而不仅仅是社会经济地位的问题。该项目将在美国中西部选定的人口普查区收集和绘制无线互联网使用地图，并将这些数据与美国人口普查的人口统计数据结合起来。虽然关于创新扩散的文献已经很发达，但对扩散的空间分析还很少见，而且有可能改变我们对新技术扩散的看法。此外，该项目将开发用于社会科学的无线测绘方法，这是一种有前途的新方法。最重要的是，该研究将根据重要的新技术和制度安排（如无许可频谱）重新考虑通信基础设施发展的理论，同时具体建立对无线数据网络的科学理解。无牌是否是管理频谱的一种可行手段，这是目前这项研究可以解决的一个紧迫的政策问题。此外，应该能够预测使用中的“漏洞”，为普遍获取领域的创新决策提供基础。例如，研究结果可能表明在经济萧条地区存在地理频谱地役权（而不是目前认为的特权空白地区）。这些经验数据还将回答一些实际问题，比如低功率无证发射机有多普遍，它们之间的距离有多近。该项目的综合教学组成部分促进了一种具有跨工程，通信和政策多学科专业知识的新型学生。将开设一门新的基于项目的课程，教学生收集无线数据。作为一个服务项目，连续几年的课程将开发一个示范互联网“地图服务器”，允许人们通过输入他们的地址来可视化、测试和改进频谱使用的预测，从而促进公众对科学和技术的理解。

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.