Research On Radio Resource Management In Heterogeneous Reconfigurable Networks

With the development of wireless communication system, radio resource management techniques are becoming crucial factors to improve the radio resource utility and meet various kinds of service requests with different QoS requirements. Future wireless communication system is commonly considered as heterogeneous wireless communication system, also known as heterogeneous wireless network. It is reported in ITU-R M.1645 recommendation that the convergence of heterogeneous wireless networks is one of the two major development directions of future wireless communication system. Software defined radio and reconfiguration technology give powerful support to the convergence and cooperation of different radio access technologies. The integration of existing radio access systems and their evolutional systems through packet based platform can realize the inter-system collaboration and interconnection providing the best access anytime anywhere. Heterogeneous wireless network can keep users always best connected and also achieve the optimal radio resource utility by pooling and managing the radio resource jointly. The radio resource management of heterogeneous reconfigurable wireless network is the key issue of heterogeneous wireless network research and attracting more and more attention.This dissertation thoroughly analyzes and studies three key technologies of radio resource management in the heterogeneous reconfigurable wireless network--joint radio resource management, dynamic spectrum allocation and dynamic network planning and management, improving the performance of current algorithms and proposing new radio resource management methods to guarantee the quality of service and increase the radio resource utility based on fuzzy logic system, graph theory, optimization theory, integer programming, genetic algorithm and other related theories and methods.First, this dissertation models and analyzes joint radio resource management of heterogeneous wireless network with queuing theory to theoretically explain the reason why joint radio resource management can guarantee the quality of service and increase the radio resource utility. Fuzzy logic algorithm is one of the most efficient ways to realize joint radio resource management. After analyzing the characteristics of different radio access technologies, a set of fuzzy rules specific to the different characteristics is designed to improve the performance of joint radio resource management based on fuzzy logic. The joint radio resource management with the radio resources reallocation function is also proposed. Through modeling the UMTS+WLAN scenario with graph theory, the radio resources reallocation method based on traffic transition is designed to provide better assurance to the quality of service and increase the radio resource utility of heterogeneous wireless network.Second, the study of dynamic spectrum allocation based on genetic algorithm is carried out. At present, most of the genetic algorithms adopted in dynamic spectrum allocation studies are standard genetic algorithm which tends to miss the global optimal solution with a specified probability. Therefore, the genetic algorithm with best reservation function is introduced, and the cell-by-cell dynamic spectrum allocation based on the genetic algorithm with best reservation is proposed to promote the performance of dynamic spectrum allocation algorithm. After analyzing the characteristics and difficulties of dynamic spectrum allocation in the urban scenario, a fast dynamic spectrum allocation method in allusion to the urban scenario is proposed. By reasonably splitting the dynamic spectrum allocation problem into several sub-problems which can be solved independently, the computational complexity of dynamic spectrum allocation algorithm is dramatically reduced. The fast dynamic spectrum allocation method can make better assurance on the real time requirement of dynamic spectrum allocation and still obtain relatively high spectrum resources utility.Finally, there is no efficient solution to dynamic network planning and management under flexible spectrum allocation so far, so research must be conducted in this particular field. The dynamic network planning and management under flexible spectrum allocation is modeled as a hierarchical optimization problem, while the inside optimization problem is modeled as the integer optimization problem and the outside optimization problem is similar to dynamic spectrum allocation problem. A hierarchical optimization method for dynamic network planning and management is designed, which integrates two kinds of optimization method--branch and bound algorithm and genetic algorithm as the solution to dynamic network planning and management under flexible spectrum allocation. The effectiveness of dynamic network planning and management under flexible spectrum allocation is demonstrated by simulation.