Author: Steven Siying Hong

Publisher:

Published: 2012

Total Pages:

ISBN-13:

Wireless has quickly become an irreplaceable medium for communication, but by their very nature, wireless transmissions are vulnerable to RF interference from various sources. This weakness is a growing problem for technologies that operate in unlicensed frequency bands, as these bands are becoming more crowded over time. The increasing density of radios, coupled with the growing number of different protocols, dictates that future wireless devices will require the ability to coexist and operate in a crowded and often unregulated electromagnetic environment. This dissertation presents a set of architectures and mechanisms for wireless coexistence. These mechanisms are intended as general primitives, enabling future wireless networks to be architected such that they are aware of who their RF neighbors are, what they are doing, and how to respond to them. We design and develop two systems, each demonstrating how changes at different levels across the network stack can provide significant gains. Degrees Of Freedom (DOF) is a robust and accurate technique to detect activity in any band of the relevant spectrum, while Picasso is an extensible radio design which allows the device to more easily exploit the fragmented bands which are available. This dissertation contributes to advancing the state of the art in wireless network design in multiple ways. First, the systems in this dissertation present novel algorithms that span signal processing, circuit design, and software systems. Not only do these algorithms enrich existing theory, they are designed with practical implementations in mind - helping to bridge the gap between theory and practice. Second, we present prototype implementations and testbed evaluations of our systems. Our results demonstrate that our systems deliver significantly improved performance in crowded interferencecentric environments, compared to the state-of-the-art. Finally, the work in this dissertation demonstrates that by having layers work cooperatively, rather than in isolation, common goals of high throughput and reliability in dense interference scenarios are much more practical to achieve. In particular, DOF demonstrates that more fine-grained information passed between the PHY and the MAC enables the MAC to make more informed and intelligent decisions regarding how to respond to the varying sources of interference. In the same vain, Picasso demonstrates that increased RF isolation can significantly lessen the burden on PHY filtering and MAC coordination - similarly allowing the three to share in the task of exploiting fragmented spectrum. We believe that the systems in this dissertation make a compelling and strong argument for cross-layer interdisciplinary wireless network design in order to solve the problems of wireless coexistence.