Research On Several Key Problems In Heterogeneous Wireless Network

Over the past decade, there has been a significant advance in the design of wireless communication networks, and a large number of wireless networks nowadays are based on different Radio Access Technologies (RAT) and standards are still being developed. Each RAT has its strengths and weaknesses with respect to capacity, cost, achievable data rates, support for quality of service and support for end user mobility. As no single RAT will be able to fully support all services and user requirements, Next Generation Heterogeneous Wireless Network (NGHWN) will be an integration of different RAT technologies and different services. Some new problems emerge in the process of convergence, such as reference model and architecture, ubiquitous mobility management, radio interface interconnection, joint radio resource management and service quality management. These issues become more and more complicated and difficult due to the heterogeneous network environment, improved user requirements and scarcity of the radio spectrum resource. Apparently, some novel theoretic and techniques are required for the solution of above issues. Since the convergence of heterogeneous networks is a huge problem which contains a lot of issues, this thesis only choose some of them as the objects for research, such as ubiquitous mobility management, network and terminal selection, cognitive radio based joint radio resource management and QoS management.The main work and contributions in this thesis can be summarized as follows.Firstly, user mobility is a typical behavior in modern mobile communication networks, and the mobility management technology has been fully studied in the existing networks. However, how to guarantee the service performance across different RATs with overlapped coverage is an urgent problem for study. The advanced mobility management is studied in this thesis and an integrated architecture has been proposed according to Next Generation Heterogeneous Communication Network architecture and international standards. Three functional modules are included in the advanced mobility management model, such as information collection, network/terminal selection and handoff decision. A novel network selection algorithm is proposed in this thesis, which integrates consideration of network/terminal performance, service and user requirements. Simulation results show that different network and terminal can be chosen for the services with different service requirements and user demands. The network selection results are shown to meet the optimization objectives of the decision maker to support ubiquitous mobility and optimal network access in the heterogeneous networks.Secondly, this thesis focuses on the cognitive radio technology in the heterogeneous networks. Since the latest popularity of wireless devices and applications led to the development of a huge variety of wireless services, the rapid increase in the demands for high-speed multimedia communications stands in clear contrast to the rather limited radio spectrum resources. Moreover, the future heterogeneous network should integrate the multiple RATs with different radio interfaces to support advanced wireless services. Cognitive radio, defined as an intelligent wireless communication system that is aware of its environment with adaptation ability and reconfigurable interoperability, is viewed as an efficient approach for improving the utilization of the precious spectrum resource and integrating the different RATs. Spectrum sensing is one of the most challenging functions in cognitive radio system. Detection of the presence of signals and distinction of the type of signals in a particular frequency band are precondition for cognitive radios to interconnect other radio interfaces in heterogeneous network. In this thesis, a novel approach to signal classification combining spectral correlation analysis and support vector machine (SVM) is introduced. Four spectral coherence characteristic parameters are chosen via spectral correlation analysis, and then optimal kernel function and parameters are selected for a nonlinear SVM via kernel alignment based method and cross-validation method. Simulations show that the proposed approach is more effective in the case of low SNR and limited training numbers compared to the existing methods.And then, a novel dynamic spectrum sharing scheme combining spectrum adaptation and MIMO-OFDM for cognitive radio system is proposed in this thesis. Given spectrum sensing and channel estimation information by the receiver, an improved model due to transmit power thresholds for SA-MIMO-OFDM system is adopted to achieve spectrum adaptation for unlicensed users. In order to improve spectrum efficiency and reduce computational complexity, a simple dynamic power allocation algorithm is also proposed. Simulations indicate that the scheme solves the partial interference problem in Interference Temperature Model and higher spectrum efficiency is achieved by multi-antenna techniques.Finally, MIMO-OFDMA and MIMO-MC-CDMA are tremendously gaining in popularity nowadays as a promising technology for supporting highspeed data transmission and low inter symbol interference in multipath fading environments. They are also prevalent candidates in future generation broadband wireless access networks, such as 3GPP Long Term Evolution (LTE), Beyond 3G (B3G) and 4G systems. The dynamic spectrum sharing scheme in multiuser MIMO-OFDMA and MIMO-MC-CDMA cognitive radio system are investigated in this thesis. In order to provide QoS support, some utility functions are designed for different services. A novel dynamic resource allocation and scheduling scheme based on user sum-utility maximization is also proposed. After being transformed into a convex optimization problem via relaxation method, the NP-hard combinatorial optimization problem can be solved by a Lagrangian primal-dual approach. The optimal solution of the primal problem can be obtained through a sub-gradient iterative algorithm. Simulation results indicate that the propositional algorithm converges to optimal power/rate allocation for primary users and cognitive radio users with maximal sum-utility and achieves optimal user scheduling with low complexity and fast convergence speed.According to above tasks and achievements, some technical papers have been published, and two technical patents have been obtained.