Study Of Spectrum Resource Utilization Based On Cognitive Radio

Cognitive radio (CR) technology has been considered as a solution to the conflict between spectrum congestion and spectrum under-utilization since it was proposed. It is applied in unlicenced systems which access the licensed spectrum bands in an opportunistic way. As an example, IEEE 802.22 system based on CR uses the broadcasting TV bands to realize the wideband spectrum access in rural area. Although the standardization work has been underway for many years, the prospect of commercial application is still unclear. Spectrum utilization problem can not be solved by unlicensed systems. Licensed systems also need to improve the spectrum utilization efficiency to solve the spectrum utilization problem.Unbalance exists in the spectrum utilization of various wireless systems. Specturm utilization of some systems is low most of the time. And spectrum of some systems is under-utilized in some time and some area. Spectrum sharing and spectrum re-allocation among different systems can reduce the unbalance in spectrum utilization. Unlicensed systems share spectrum resource with licensed system in an opportunistic way. This method can also be applied in other systems to solve the spectrum resource allocation problem in heterogeneous network environment.Cognitive radio systems should discover the available spectrum resource before using it. The effect of spectrum utilization is closely related to the spectrum detection performance, and spectrum access strategy, both of which are discussed in this dissertation. The major research contents consist of the following parts:The concept of spectrum resource is illustrated. Different ways of using spectrum resource of various wireless technologies is concluded in the perspective of spectrum space. Based on the spectrum resource space model, the situations of unlicensed system and licensed system are explained, the differences between these two situations are also pointed out. A general framework of evaluation criterion of spectrum utilization is proposed.The performance of different detection methods and the influence of the noise uncertainty are analyzed. Discrete and continous forms of noise uncerntainty model are proposed. The design criterion is given using average detection probability and average false alarm probability as the indictors to evaluate the detection performance. Compared with traditional research in the detection under noise uncertainty, the model proposed in this thesis is more accurate. If the threshold of the detector is determined according to the prior knowledge of noise variance distribution, the average detection performance will be achieved.The cooperative spectrum sensing issue is studied in the perspective of multi-sensor data fusion. Neglecting the detection result of some unreliable users, the detection performance can be improved by allowing part of the users participated in data fusion. For a certain amount of online users, there exsits an optimal number of users participated in data fusion. Selecting a fixed number of users participated in data fusion as the optimal value can maximize the average system detection probability. Dynamically adjusting the number of users participated in data fusion by measurement can maximize the detection performance in each detection.Terminal-dominated spectrum allocation scheme is proposed on the fundamental of multi-mode terminal and cognitive radio technology. From the view of network operators, although each user doesn't know the situation of other users in this scheme, high effiency of spectrum utilization of this scheme will be achieved with proper access strategy. From the view of users, the automatic tradeoff among heterogenous network of multi-mode terminal is helpful to ensure the QoS of application for the users in a mobile environment.The requirements for the next generation multi-mode terminal are described. The prospect and technological viability of multi-mode terminal is analyzed. The architecture and operating mechanism of CR-based multi-mode terminal is proposed to support the terminal-dominated spectrum allocation scheme. Enabling technologies needed by the intelligent terminal are discussed.Modeling method and design of simulator of cognitive radio system are proposed. The simulator is based on object-oriented design thought, which is benefit for the extention and adaptation of the functions of the simulator. The simulator applies discrete event driven mechanism to emulate the process of system operating. Channel is the focus in the simulator, and the situation of all the transceivers can be directly observed all the time, which makes it easier for the evaluation of spectrum resource utilization.