Research On Multi-channel Optimization Strategy In Wireless Mesh Network

With the development of wireless communication technology, wireless networks that provide flexible Internet access emerge in our daily life. The widely used wireless local network (WLAN) is a one-hop network, which provides the last one mile Internet access. Because the one-hop wireless communication range is limited, it's very dependent on the wired networks to cover a large area. In order to meet the needs of large area and high bandwidth Internet access, wireless mesh networks (WMNs) emerged.WMNs provide Internet access through multi-hops. As a revolutionary innovation of network topology, WMNs expand the coverage range, reduce the cost of deployment, and provide high bandwidth. WMNs are in research stage now, and there is no unified standard for WMNs.Wireless signals will interfere with each other if they have the same frequency. If two or more wireless devices work on the same channel, they will share the bandwidth. So, the multi-channel technique can increase the capacity of WMNs. In order to take full advantage of multi-channel technique, wireless nodes in WMNs often have multi-radios. How to use multi-channel technique to improve WMN performance is one of hot research topics, and that's what this thesis focuses on.Firstly, a wireless mesh network testbed is deployed to study the transmission characteristic of WMNs in real environment. In order to evaluate the performance of protocols in WMNs, wireless mesh network modules are added into NS2.Secondly, the interference between different channels is investigated in this thesis. It consists of two parts, experiment and theoretical analysis. Through the experiment on the testbed, we evaluated the channel interference between links in same path and different paths. Then, we made some recommendations to channel selection. Through the theoretical analysis, we proposed the metric of the channel to links and nodes for channel assignment.In IEEE 802.11s draft standard, the channel assignment algorithm is not completed. Therefore, two channel assignment algorithms in two ways are proposed in this thesis. One is a minimize interference channel assignment algorithm, the other is based on link quality. Both of them are distributed and dynamic channel assignment algorithm, with high flexibility. The simulation results show that the channel assignment algorithms we proposed in this thesis have higher performance than conventional channel assignment algorithms.