| Summary: | 碩士 === 元智大學 === 電機工程學系 === 99 === In cognitive radio (CR) networks, available channels of each secondary user (SU) is dynamic and time-variant, thus a flexible approach which lets SUs connect with each other is needed. In this paper, we base on an existing framework for ad hoc networks, and propose two types of cluster-constructing methods which consider connect probability. The first method called probability-based (PB), by accurately calculating each node''s expected connectivity, and it determines each node''s priority of becoming clusterhead (CH). The other method call approximated probability-based (APB), which considers the limitation of physical antennas, and calculates the approximated connectivity. We also perform simulations to compare the performance between the proposed methods and the tradition one. The simulation results show that, in different conditions (ex. SU density, PU density, transmission range, channel number etc.,) the performances of PB and APB are always better than the traditional one, which indicates that the two proposed methods can enhance network performance by considering the connectivity.
|
|---|
