MODELS AND ALGORITHMS FOR SPECTRUM COEXISTENCE IN WIRELESS NETWORKS Models and Algorithms for Spectrum Coexistence in Wireless Networks
In the last decade, there have been several technological trends that have occurred together and have caused a shift in how wireless systems will be deployed. The significant increase in the capabilities of mobile devices, combined with the proliferation of Internet enabled services, and the improvement in the communication support provided by new waveforms for wireless communications, have initiated a shift from the traditional, macrocell-based cellular network to new forms of radio access
... of radio access technologies (RATs) involving multiple, smaller cells deployed in vicinity of each other. These small cells will often support diverse wireless technologies and be operated by different providers. The resulting heterogeneity, unfortunately, can lead to serious internetwork interference that can negate the improvement in overall system performance that was the original motivation for employing many small cells in close proximity. In this thesis, we examine different technologies that are needed for flexible spectrum management to support the coordination that is needed for coexistence between many small cell wireless networks. Motivated by the need for internetwork architectures that support spectrum coordination, we (1) conduct performance evaluation associated with the joint ii deployment of mobile and fixed hotspot networks, (2) develop spectrum models that characterize interference among different wireless entities, (3) provide new methods for efficient hardware emulation of wireless channels, (4) devise algorithms that estimate radio spectrum usage, and (5) provide algorithms for coordination between different wireless systems to improve the overall system performance and spectrum efficiency. The first part of the thesis investigates spectrum coexistence in wireless networks by exploring the underlying performance challenges that exist when mobile hotspots are deployed in an environment of densely deployed, static wireless access networks. Next part of thesis investigates design of hardware emulator of radio channels to accurately capture the effect of real-world wireless channels upon communications waveforms while minimizing computational complexity. Next, we explore a fundamental building block of spectrum management for supporting better utilization of radio spectrum which involves predicting the impact that an emitter will have at different geographic locations. We then examine various challenges associated with coordinating spectrum access between different wireless technologies by exploring the specific case of Wi-Fi and LTE coexistence in emerging unlicensed frequency bands. Finally, recognizing the broad challenges associated with addressing spectrum coexistence in emerging wireless systems, we identify several directions for future investigation and suggest different approaches for tackling these challenges. iii Acknowledgements "Life is a journey, not a destination." − Ralph Waldo Emerson My graduate school journey over the last six years could not have been described more aptly than this favorite quote of mine. When I landed in the US for my master's, my plan was to do master's and take a job. But then WINLAB happened which changed me fundamentally in the way I function and think, for which I always be grateful. I feel fortunate to be guided by Prof. Wade Trappe, Prof. Dipankar Raychaudhuri and Dr. Larry Greenstein, whose working style though differ from each other, made my experience a wholesome with their technical and life advice and constant support. Prof. Ray has always inspired me with his simplicity of logic, heuristic approach and acute engineering insights. Wade drove my thought process to new ideas that I can immediately connect with. I never failed to be amazed how Larry's casual technical intuitions are always close to the profound theory.