Studies On Temporal-frequency-spatial Three-dimensional Spectrum Hole Based Opportunistic Access In Cognitive Radio Networks

In cognitive radio networks,in order to tackle the problems of spectrum scarcity and underutilization, the secondary users can utilize the spectrum holes to access the network without causing harmful interference to the primary system. One of the key issues in cognitive radio is the effective detection and access to the spectrum holes. However, the conventional opportunistic access schemes have not fully utilized the whole spectrum hole, by restricting only to temporal and frequency dimensions.Thus, in this dissertation, the opportunities in temporal,frequency and spatial dimensions are explored with mathematical modeling and quantitative analysis.The features of spatial spectrum holes and temporal-frequency spectrum holes are also differentiated. Furthermore, the opportunistic access problems have been investigated based on the utilizations of spectrum holes with three dimensions, for both small-scale and large-scale cognitive radio networks.In small-scale cognitive radio networks,the spectrum opportunities surrounding the secondary users exhibit great similarities due to the node adjacency. As a result, the spatial spectrum holes are very limited and the secondary users should seek to an effective exploration of temporal and frequency opportunities.In this dissertation, a distributed opportunistic access scheme is proposed, which allows the secondary users to search for the access opportunities in multiple channels.Furthermore, the effective throughput is also optimized. To further improve the spectrum utilization, a cooperative access scheme is then proposed and the upper and lower bounds on the cooperative throughput are also calculated. The proposed scheme utilizes the spectrum holes efficiently and thus is able to yield a near-optimal system throughput. In addition, fairness is an important issue on characterizing system operation efficiency and user satisfaction.Hence, the fairness issues are also analyzed when multiple secondary users compete to access the spectrum.In particular, a fast catch-up strategy is then proposed to improve the fairness of the original algorithm.In large-scale cognitive radio networks, the different spectrum environments surrounding the secondary users create vast spatial potentials for access.The effective utilization of the spatial spectrum holes relies heavily on transmit power control, where the knowledge of the position information of the primary users becomes instrumental.However, the conventional localization algorithms for wireless networks are not applicable to localize the primary users.Accordingly, this dissertation proposes a high-order range-free localization algorithm, which improves the localization accuracy of range-free algorithm by taking advantage of much more surrounding secondary users.Furthermore, a semi range-based localization algorithm is proposed, by virtue of the estimation of the average detection probabilities through the binary detection results.This algorithm is then extended to the 3-dimensional localization so as to fulfill more practical requirments.More importantly, analytical and simulation results show that this approach performs closely to the Cramer-Rao lower bound and thus provides excellent localization accuracy. Finally, we shall propose an opportunistic spectrum access scheme with channel selection and power control mechanisms for distributed and centralized networks,respectively, where localization errors are also considered in the scheme design.The proposed algorithm is proved to greatly enhance the performance of the secondary system by effectively utilizing the spatial opportunities.