The objective of this research is to investigate the power-control problem and analyze the network capacity in cognitive radio (CR) networks. For CR users or Secondary users (SUs), two spectrum-access schemes exist: namely, spectrum underlay and spectrum overlay. Spectrum overlay improves the spectrum utilization by granting SUs the authority to sense and explore the unused spectrum bands provided by PUs. in this scheme, designing effective spectrum-sensing techniques in PHY layer is the major concern. Spectrum underlay permits Sus to share the same spectrum bands with PUS at the same time and location. In this scheme, designing robust power control algorithms that guarantee the QoS of both primary and secondary transmissions is the main task. In this thesis, we first investigate the power-control problems in CR networks. Especially, we conduct two research works on power control for CDMA and OFDMA CR networks. Being aware of the competitive spectrum-access feature of SUs, the non-cooperative game theory, as a standard mathematics, is used to study the power-control problem. Note that game-theoretical approaches provide distributed solutions for CR networks,, which fits the needs of CR networks. However, it requires channel state information (CSI) exchange among all SUs, which will cause great overheads in the large network deployment. To gain better network scalability and design more robust power-control algorithm for any hostile radio-access environments, we propose a reinforcement-learning-based repeated power-control game that solve the problem for the first time. The left part of the dissertation is to study the throughput capacity scaling of the newly arising cognitive ad hoc networks (CRAHNs). Stimulated by the seminal work of Gupta and Kumar, the fundamental throughput scaling law for large-scale wireless ad hoc networks has become an active research topic. This research is of great theoretical value for wireless ad hoc networks. Our proposed research studies it in the scenario of CRAHNs under the impact of PU activity. It is a typical and important network scenario that has never been studied yet. We do believe this research has its unique value, it will have an impact to the research community.
| Identifer | oai:union.ndltd.org:GATECH/oai:smartech.gatech.edu:1853/44736 |
| Date | 16 May 2011 |
| Creators | Zhou, Pan |
| Publisher | Georgia Institute of Technology |
| Source Sets | Georgia Tech Electronic Thesis and Dissertation Archive |
| Detected Language | English |
| Type | Dissertation |
Page generated in 0.0168 seconds