AbstractsPhysics

Dynamic Spectrum Access (DSA) is a new spectrum sharing paradigm that aims at increasing the utilization of radio frequencies as well as mitigating the spectrum scarcity problem. It allows Secondary Users (SUs) to access idle channels in the licensed spectrum while protecting the signals of Primary Users (PUs) from harmful interference. Although local-sensing can detect an empty radio frequency band, it does not provide a sufficient level of protection to a PU's transmission. Therefore, DSA currently relies on online geo-location databases, White Space Databases (WSDBs), to distribute information about the availability of radio spectrum white spaces. In this thesis, we make three contributions to the field of Dynamic Spectrum Access (DSA). First, the access to WSDBs in terms of response message size and response time is profiled using the state-of-the-art WSDB query technique it{Single-location WSDB query}. This comparative study helps us in understanding the differences between the WSDBs' performances and shows us the capabilities and limitations of the current query technique. Our conclusion from this part of the study is that the current WSDB query method has poor support to white space-enabled mobile devices (MWSD). Second, based on our previous conclusion, we propose a new WSDB querying technique it{multi-location WSDB query} that optimizes the access of MWSD to a WSDB by reducing the number of needed queries to cover a particular path and as a results the energy consumption of the query process is reduced. Furthermore, We show that it{multi-location query} has a much quicker response time than it{single-location query}. We introduced the world's first algorithm called it{Nuna} that predicts the next direction of a movement path, estimates the required size of the multi-location WSDB query, and queries a WSDB. it{Nuna} helps MWSD to use it{multi-location query} without they need to know their path in advance. Moreover, we demonstrated by example that it{Nuna} halves the required number of queries to cover a particular path, improving the energy efficiency by 50%. Finally, we shifted our focus from optimizing the access of MWSD to a WSDB to optimize the information of a WSDB by providing local sensing readings to a WSDB. Therefore, we developed an energy efficient Portable Spectrum Sensing Platform (PoSSP) that can obtain its GPS position, sense the spectrum, and send its data to an online server. We show that PoSSP is by 60% more energy efficient than similarly designed sensing platforms. Advisors/Committee Members: Przemek, P..