Investigation Of Network Control And Radio Resource Management In Metropolitan Area Wireless Mesh Networks

Providing users with high-speed Internet access and QoS-guaranteed multi-service broadband connection in a flexible manner is the target of next generation wireless communication networks. Wireless mesh networks adopt multi-hop self-organized networking model, extend coverage of wireless metropolitan area networks, support both fixed broadband Internet access service and peer-to-peer traffic transmission, combine centralized and distributed network control and radio resource management, and thus overcome the un-controllerable and unsafety problem of ad hoc networks. Metropolitan area wireless mesh networks are undergoing commercialization as the first stage, which have attracted increasing interest in the theoretical research of wireless networks, while related standards and amendments are being developed at the same time.In this thesis, interference models for IEEE 802.16 standards based wireless mesh networks under condition of single channel and multi-channel are established. Based on these models, network capacity in different forms and under various conditions is analyzed. And a series of network control and radio resource management algorithms are investigated for the case of single channel and multi-channel, omnidirectional antenna and directional antenna separately. Theoretical analysis and system-level dynamic simulation results are also provided.Firstly, main results of network capacity analysis for distributed, centralized, and hybrid single-channel wireless mesh networks are summarized. Construction method of getting interference model for single-channel wireless mesh networks from node topology is described. Different forms of this model are discussed under different antenna conditions. The relationship between network capacity and node numbers of different levels for single channel wireless mesh networks with regular binary tree topology is analyzed. And the relationship between network capacity, route distance and hop count for single channel wireless mesh networks with irregular chain topology is also analyzed.Secondly, network control and radio resource management problems, such as routing, scheduling, and power control for single-channel wireless mesh networks are summarized. The system framework for single channel wireless mesh networks is modeled, including network architecture model, link interference model, tree-type routing model, as well as uplink and downlink traffic model. A maximal clique based fair downlink scheduling algorithm and a tree-type dynamic programming based uplink scheduling algorithm are proposed for the case of omni-directional antennas. A tree level based downlink scheduling algorithm and a buffer size and link condition based uplink scheduling algorithm are proposed for the case of directional antennas. Moreover, both iteration method based and game theory based distributed power control algorithms are proposed under condition of omni-directional antennas.Thirdly, three typical methods of network capacity analysis for multi-channel wireless mesh networks are summarized. Construction method of getting multi-radio multi-channel (MRMC) interference model from node topology and channel and radio constraints is described. The relationship between network capacity of MRMC wireless mesh networks and some kind of ratio between channel numbers and radio numbers is obtained by applying information theory based capacity analysis method. Furthermore, the critical point between the radio constraint scope and the channel constraint scope and its affection factors are also obtained by applying MRMC interference model based capacity analysis method.Finally, three main route construction and channel allocation algorithms for multi-channel wireless mesh networks are summarized. The system framework for MRMC wireless mesh networks is modeled, including network architecture model, multi-channel link interference model, tree-type routing model, as well as uplink and downlink traffic model. Also, an uplink and downlink mixed scheduling algorithm based on sub-optimal solutions and a joint routing and scheduling algorithm based on linear programming optimizations are proposed for MRMC wireless mesh networks.