Resource Allocation In Cooperative And Multi-channel Cognitive Radio Networks

The explosive growth of wireless traffic, smart devices, and new applications will intensify the need for spectrum resource. However, investigation shows that the current static spectrum access policy fails to utilize the spectrum resource efficiently and the "spectrum hole" exists at certain time and space. In parallel with that, the proposed cognitive radio technology allows secondary users to sense transmission environment and dynamically access the spectrum resource with limited interference to primary users. As a result, the cognitive radio could greatly increase the spectrum utilization and improve the network performance. The resource allocation problem is widely investigated in cognitive radio networks, including how to design spectrum sensing strategy, how to utilize and allocate network resource, and how to manage the interference among users. In this paper, we focus on the resource allocation problem in three cognitive radio networks. Specifically, we first study the joint relay selection, secondary transmission period allocation, and power allocation problem in central-ized single-hop cooperative cognitive radio networks. Then, we investigate the joint channel allocation, radio assignment, and flow routing problem in centralized multi-hop multi-channel multi-radio cognitive radio networks. Finally, we extend the analysis to the cognitive radio networks without centralized infrastructures, and design the dis-tributed resource allocation algorithm in multi-hop multi-channel multi-radio cognitive radio networks.The main contributions of this dissertation could be summarized as follows:· In the centralized single-hop cooperative cognitive radio networks, we study the resource allocation problem by jointly considering relay selection, secondary transmission scheduling, and secondary power allocation. We allow all sec-ondary users to utilize the cooperation-generated resource. This not only im-proves the spectrum utilization, but also introduces the user diversity to increase network throughput performance. Considering the energy limitation on sec-ondary users, we maximize the long-term averaged secondary network through-put, with primary performance constraint, network stability constraint, and sec-ondary energy constraint. Through Lyapunov theory, we design an online al-gorithm to solve the long-term network throughput maximization problem. In the algorithm, we first study the simple case with given power, and then extend the results to the complicated case by considering power allocation. Finally, the complexity is also analyzed for the proposed algorithm.· We study the centralized resource allocation algorithm in multi-hop multi-channel multi-radio cognitive radio networks. We propose a novel conflict graph based on link-channel-radio pair, to describe the relationship between radio re-source and channel resource. Based on this conflict graph, we prove the suf-ficient condition for interference-free scheduling. With this sufficient condition, the complexity of searching all the maximal independent sets in conflict graphs could be avoided. Moreover, considering the characteristics of scalable video coding streams, the flow routing, channel allocation and radio assignment are jointly optimized, to maximize the network throughput under fairness constraint.· Based on the analysis in the above centralized networks, we study the distribut-ed resource allocation algorithm in multi-hop multi-channel multi-radio cognitive radio networks. Under Network Utility Maximization theory, the convex optimiza-tion problem is solved through its dual problem. Then we decompose the dual problem into multiple sub-problems, each of which can be distributedly and lo-cally solved through the secondary users and links. Furthermore, we exploit Proximal Optimization algorithm to transform the non-strictly concave optimiza-tion problem into an equivalent strictly concave problem, to avoid the oscillation problem caused by multi-path flow routing. Finally, the algorithm performance is evaluated through extensive numerical analysis.