Research On Dynamic Spectrum Allocation Technology In Cognitive Wireless Networks

Recent developments of wireless communications need more radio resources to provide services with high quality and data rate, which leads to the scarcity of the public radio spectrum being a serious problem. With the dramatic growth of communication applications, the precious spectrum resource becomes more and more congested. Recent survey has in fact proved that most of the radio frequency spectrum is vastly under-utilized in both the temporal domain and the spatial domain. To deal with the inconsistency between spectrum congestion and spectrum under-utilization, many recent researches have focused on Dynamic Spectrum Allocation (DSA) technique.With the evolution of radio technology, the concept of end-to-end reconfiguration enables the mobile terminals to change their frequencies dynamically so as to access different spectrum bands agilely. Moreover, Cognitive Radio (CR) technique has been proposed to improve the spectrum utilization efficiency dramatically by allowing non-legitimate users to sense and access vacant licensed bands opportunistically. Cognitive Wireless Networks (CWN) based on cognitive radio exploit cognition principles in wireless networks to improve network performance and utilization efficiency of radio resources. Dynamic spectrum allocation technique is applied in cognitive wireless network to increase spectrum utilization efficiency.Dynamic spectrum allocation is a key technology in cognitive wireless network related researches. This thesis works on DSA technology, including system architechture, algorithms and mechanisms design for different scenarios. DSA enables the dynamic allocation of spectrum resource to dig spectrum holes adequately, so as to maximize the efficiency of spectrum utilization and improve network performances.In general, the contents of this thesis include following aspects:Firstly, the development trend of future wireless communications networks is envisaged. Efficient Spectrum utilization and heterogeneous network convergence is identified as an inevitable trend in future wireless communications field. The SDR, CR and AI technologies are introduced, including the origins, present status and future development. Moreover, the fundamental of cognitive wireless network and the key problems are introduced, including cognition, CPC and so on. The DSA technology is analyzed in detail, including its research areas, significance and present status.To implement the DSA technology, the wireless network architecture as a support is necessary. This thesis designs a novel cognitive wireless network architecture based on DSA, including the definition of function module and interface between the modules, as well as the key function procedure of signaling exchange. In this architecture, SDSA and TDSA mechanisms are proposed. The configuration scheme and the cooperation scheme of function mudule are also designed. Moreover, the mapping schemes of function architecture in different application scenarios are brought forward. The advantages and disadvantages of the mapping schemes are discussed indetail to provide the reference for the deployment of actual networks.As for short-term dynamic spectrum licensing, a group buying based dynamic spectrum allocation (GB-DSA) mechanism is brought forward to enhance the spectrum utilization efficiency and maximize the profit of primary systems by adjusting pricing dynamically in cognitive wireless networks. In parallel with this, the in-group pricing allocation mechanism is established based on the individual cell's utility contribution. As a consequence, each secondary buyer in the group could achieve better performance with lower rate owing to the contribution of group and the utility of the whole secondary systems will be improved effectively. Especially, an interference mitigation method is brought forward to solve the inter-system interference problem, which is applied as a fundamental issue in constructing the "Group". In this scheme, game theory and optimization theory are applied in the GB-DSA scheme to investigate a win-win solution for spectrum trading.Due to the dynamic and random characteristic of radio environment in cognitive wireless network, this paper brings forward a dynamic channel allocation mechanism to minimize the probability of collision with the licensed users, eliminate the phenomenon of overload, reduce blocking probability and improve the performance of cognitive wireless network. Matching rules are set forth by considering relationship between spectrum statistic characteristic and service characteristic adequately. Furthermore, a multi-objective terminal selection algorithm is designed to minimize the system overhead and the impact on end-users.As for intra-network cooperative dynamic spectrum sharing, two new dynamic spectrum access schemes for secondary users are proposed to guarantee fairness, which are based on continuous-time Markov chains (CTMC). Both the interactions between primary and secondary users and the effects of imperfect spectrum sensing (i.e. miss-detection and false alarm) are considered in the schemes. Due to the fact that miss-detection would lead to collision between primary and secondary users, we derive the optimal access probabilities for secondary users, so that the QoS of primary users in terms of collision probability is guaranteed.A summary is given at the end, where the future research directions related to this thesis are also pointed out.