Research On Key Technologies Of Resource Management In Multi-Radio Multi-Channel Wireless Mesh Network

Wireless mesh network (WMN), which is a new broadband wireless network structure, consists of mesh routers and mesh clients, where mesh routers have minimal mobility and form the wireless backbone network to provide network access for mesh clients. WMN has many merits, such as fast construction, low upfront investment, bestrow scope, good scalability, etc. WMN usually provides internet access for the areas with no optical fibre as a wireless backhaul network, provides extensiveness internet access as a WMAN, and can also apply to emergency response administration, military or police communication systems. WMN becomes the research focus for its broad prospect of application.Channel, rate, power, time slot and code are called as radio resource. Radio resource management is to allocate radio resource neatly and adjust the allocation dynamically under unbalance load, fading and interference, to guarantee the quality of service and improve resource utilization.In traditional WMN with single radio and single channel configuration, the network capacity decreases with the increasing of the number of mesh routers. The multi-radio and multi-channel technology can decrease interference and increase network capacity greatly. However, the complexity of resource management also increases with multi-radio and multi-channel technology. The effective radio resource management for multi-radio and multi-channel wireless mesh networks (MRMC-WMN) is an urgent task, the aim is to improve the network performance and quantity of service (QoS). In this dissertation, we focus mostly on the methods and strategies of radio resource management in MRMC-WMN, involves the key technologies as channel allocation, rate assignment, routing, congestion control, power control, QoS guarantee, etc.Cross layer design can optimize network performance through joint design approaches, so that the network can provide better QoS with dynamically changing parameters and limited resources. Based on the economics theory, network resources can be managed as commodities by using the achievements of economics theory. In this dissertation, we use resource management technologies based on cross layer and economics theory to study the key technologies of radio resource management jointly in MRMC-WMN. The main contributions are listed as follows:1) With the method of cross layer resource management, and based on protocol interference model, we study the resource management problems in MRMC-WMN as follows: a) Model the cross layer joint channel allocation, rate assignment and routing problem as a mixed integer linear programming problem with max-min fairness. A tabu search based iterated local search (TSILS) algorithm is proposed to find the approximate optimal solution. Compared with existing algorithms, the TSILS algorithm has better convergence and less optimal bias.b) Extend the work of a), model the joint congestion control, channel allocation, rate assignment and routing problem as a network utility maximum (NUM) problem. Because the centralized optimal algorithm has high complexity and signaling overhead, a cross layer optimization mechanism which is feasible for practical system is proposed based on convex optimization theory. In the mechanism, each layer in the protocol and each node can interchange information, so the adjustment of rate for flows in the transport layer, rate assignment and scheduling in the data link layer, and channel allocation in the gateway are coordinated. The mechanism can convergence rapidly and guarantee fairness among flows.c) Extend the works of a) and b) further, we study how to guarantee the QoS of video streams in MRMC-WMN. The channel allocation, rate assignment, delay partitioning, distortion are modeled. Because the centralized optimal algorithm has high complexity and signaling overhead, a cross layer optimization mechanism which is feasible for practical system is proposed based on convex optimization theory. The proposed algorithm can convergence to approximate optimal solution. The influence of the number of radios and channels, and the delay on distortion and rate of video streams is also analyzed.2) With the method of cross layer resource management, and based on physical interference model, in order to optimize network utility of MRMC-WMN, the joint congestion control, channel allocation and power control problem is studied. Because the centralized optimal algorithm has high complexity and signaling overhead, a distributed near optimal-algorithm is proposed based on convex optimization theory. In the proposed algorithm, source node of each flow adjusts its flow rate distributedly, each node allocates channels to its responsible logical links distributedly, and each transmitter adjusts its transmit power distributedly. The proposed algorithm can convergence to near optimal solution, and has lower complexity compared to centralized optimal algorithm, so it is feasible in practical MRMC-WMN.3) With the resource management method based on economics theory, in order to optimize network utility of MRMC-WMN, the joint channel allocation and power control problem is studied. Because the centralized optimal algorithm has high complexity and signaling overhead, a power-interference pricing model is proposed based on economics theory. Channel assignment and power allocation are coordinated through power prices and interference prices. A distributed jointly channel assignment and power allocation algorithm based on pricing with low complexity is proposed for MRMC-WMN. The proposed algorithm uses power prices to encourage nodes adjust power of links according to power demand, and uses interference prices to encourage nodes allocate channel and power rationally. The proposed algorithm can convergence to near-optimal solution faster and steadier than the existing algorithm.