A Study On Resource Allocation In OFDM-Baed Cognitive Ra-dio Networks

In the field of wireless communications, spectrum scarcity is the main obstacle of the development of wireless services. On the other hand, Federal Communica-tion Commission(FCC) reports show that the licensed spectrum is actually far under-utilized. Cognitive radio (CR) is recognized as a highly promising technology to im-prove the spectrum efficiency and gained more and more attention in recent years. In a CR network, CR users are allowed to opportunistically access to the licensed spectrum which is not occupied by the PUs at that time, as long as the interference introduced to the PUs is under a predefined threshold. Orthogonal frequency division multiplexing (OFDM), which offers high spectrum efficiency and flexibility in resource allocation (RA), has been widely deemed as a potential air interference for CR networks. For OFDM-based CR networks, adaptive RA is one of the most important issues as high performance is expected. In this thesis, we study the RA problem in OFDM-based CR networks. The main contributions of this thesis are summarized as follows.1. We developed a general models to illustrate the RA problem in the arising OFDM-based CR networks, in which both the power budget of CR network and inter-ference limit to each PU are both taken into consideration. Compared to the models in other literatures, our models are more practical for CR networks.2. We analyzed the RA problem extensively and proposed an efficient fast barrier method. The remarkable characteristic of our proposed algorithm lies in the fact that it always works out the optimal solutions with almost linear complexity by exploiting the structure of the formulated optimization problem. Compared with other algorithms in the literature, which usually adopt heuristic methods to obtain suboptimal solutions, our method are more attractive. 3. We proposed an efficient scheme for resource allocation in multi-user CR net-work with consideration of fairness among SUs, which is realized in two steps. First-ly, a proportional fair algorithm for subcarrier allocation is applied. Nextly, power is distributed by using the fast barrier method while maintaining the proportional rate constraints. Simulation results show that the proposed algorithm exhibits a good trade-off between sum capacity maximization and proportional fairness. Furthermore, it has lower complexity compared with other algorithms.4. We investigated RA in an OFDM-based CR network with cooperative relay and formulate the corresponding optimization problem. The fast barrier method is applied within an alternating optimization framework. Our model is general and the alternating framework shows some light on tackling the power allocation problem efficiently in CR network with cooperative relay.In summary, our proposed algorithms perform well with lower complexity than the other methods in literature and our research results throw some lights on design effective and efficient algorithm for the RA problem in OFDM-based CR networks. Parts of them have been published or submitted in leading journals and conferences in the related fields.