Research On Dynamic Resource Allocation In Cognitive Radio Networks

The number of wireless systems and services has grown exponentially over time, while the spectrum resource is exhausted. How to make the limited spectrum support the increasing bandwidth requirement continually is a problem to be settled. Cognitive radio can improve spectrum utilization effectively through sensing and accessing the licensed spectrum. In cognitive radio network, the key problem is to allocate the resource of each cognitive node without harmful interference to the primary users. In this thesis, we make a research on the problem of allocating network resource optimally, e.g., caching, transmit power, bandwidth, from the view of network architecture, user access control and user satisfaction. The main results are as follows:1. Research the scalability problem of cognitive radio network architecture. Due to the difference of space-time and geographical location, the state information of each cognitive user is dynamic and independent. As a result, the communication between users, especially multi-hop relaying, has high overhead. We propose a software defined cognitive radio network architecture with separation of control and forward, and design a caching placement scheme for the network.2. Research the power control in single hop cognitive radio networks. We aim to minimize the total energy and maximize the number of access users at the same time. Considerable literature has studied how to remove a subset of the secondary users such that the remaining users satisfy the constraints with power budget and Signal-to-Interference-plus-Noise Ratio. However, the removal algorithms take long interference time when the outage probability is large, i.e., there are numerous unsupported secondary users in the network. We propose a hybrid spectrum access algorithm to promote the capacity and save the interference time.3. Research the optimal allocation problem in multi-hop networks. We propose an efficient spectrum allocation algorithm under physical interference model. To reduce the computation time, we first consider pair-wise interference to construct a 4-dimensional conflict graph(4-DCG) for channels, power levels, links and radio interfaces, and find a set of maximal independent sets based on 4-DCG. We formulate the bandwidth allocation problem as an optimization problem. Since every router has caching capabilities, the network may have different traffic patterns. We propose different heuristic algorithms for various kinds of traffics, including, multi-cast, converge and hybrid.In summary, we study the network architecture and resource allocation problem in cognitive radio network. The main objective is to improve network efficiency and user experience. The research can provide theoretical guidance for the design and implementation of future cognitive radio networks in practice.