Date of Award
Doctor of Philosophy (PhD)
Linda L. Vahala
Frederic D. McKenzie
The goal of this dissertation is to present the analysis and optimization of dynamic spectrum sharing for cognitive radio networks (CRNs). Spectrum scarcity is a well known problem at present. In order to deal with this problem, dynamic spectrum sharing (DSS) was proposed. DSS is a technique where cognitive radio networks dynamically and opportunistically share the channels with primary users. The major contribution of this dissertation is in analyzing the problem of dynamic spectrum sharing under different scenarios and developing optimal solutions for these scenarios. In the first scenario, a contention based dynamic spectrum sharing model is considered and its throughput analysis is presented. One of the applications of this throughput analysis is in finding the optimal number of secondary users in such a scenario. The problem is studied for fixed and random allocation of channels to primary users while secondary users try to opportunistically use these channels. Primary users contend for the channels, and secondary users try to use the channels only when primary users are not using it. These secondary users themselves contend for the opportunistic usage. The numerical formulas developed for finding the optimal number of secondary users have been carefully analyzed with the solutions obtained using the throughput model directly and finding the optimal number of secondary users. These two match very closely with each other and hence provide simple numerical formulas to calculate the optimal number. The second scenario studied is based upon the idea of pre-knowledge of primary user activity. For instance, the active broadcasting periods of TV channels can be obtained from past measurements as the TV channels activities are approximately fixed. In this scenario, time spectrum block (TSB) allocation for DSS is studied. Optimal TSB allocation is considered to minimize the total interference of the system and hence maximize the overall throughput of the system of community networks. The results obtained using the proposed ABCD algorithm follow very closely with the optimal results. Thus the simple algorithm developed can be used for time spectrum block allocation in practical scenarios.
"Analysis and Optimization of Dynamic Spectrum Sharing for Cognitive Radio Networks"
(2010). Doctor of Philosophy (PhD), dissertation, Electrical/Computer Engineering, Old Dominion University, DOI: 10.25777/7dk9-1p88