| Summary: | 碩士 === 國立成功大學 === 資訊工程學系碩博士班 === 96 === With the development and growing of wireless transmission techniques and handheld equipments, the system adapting mobile environment is an important research issue recently. In traditional researches in static wireless networks, we use specific mathematical channel models to model the channel noise, and then optimize the parameters. For example, a common way to model the transmission channel is to use Gaussian noise model. However, in mobile wireless networks, users are always moving so that the transmission channels between users and the base station are altering all the time. Moreover, transmission channels among users are also difficult to predict. Thus, it is hard to adopt optimization process by designing specific objective functions and constraints. Besides, constantly detecting channel models is infeasible in mobile environment.
To solve the problem, this thesis provides a fuzzy feedback control system model. Because a fuzzy controller only refers to the system feedback, we do not have to design objective functions and constraints beforehand. This characteristic makes a fuzzy controller totally apply to mobile wireless networks. Furthermore, we design the system model as a hierarchical model compromising the cross-layer multi-objective problem, called hierarchical cross-layer fuzzy control (HCLFC). The hierarchical architecture not only makes the system perfectly compatible in the open systems interconnection (OSI) protocol layers but greatly reduces the computing complexity. At the same time, we adopt fuzzy decision making method to compromise multi-objective control satisfying multiple control targets with highest satisfaction among different protocol layers in wireless networks. Eventually, we adopt HCLFC for collaboration of physical (PHY) and application (APP) layers where physical layer modulation and application layer packet adaptation are compromised to meet traffic requirements. The PHY-APP controller is implemented on inter-vehicle communication (IVC) systems providing a random and rapid changing transmission channels. By the architecture design, analysis and simulations, the HCLFC model in this thesis, which is soft computing, is a better choice to accommodate to highly uncertain mobile wireless networks than the traditional optimization method, which is hard computing.
|
|---|
