Cross-layer Design Of Routing And Channel Assignment In Cognitive Wireless Mesh Networks

The rapid development in new communication technologies of wireless broadband networks becomes a challenge to the problems of scarce spectrum resources and low average utilization rate. Cognitive Wireless Mesh Networks is the combination of cognitive radio technology which is using dynamic spectrum management to improve the spectrum utilization and Wireless Mesh Networks which is the most promising next generation wireless network technology. It can achieve heterogeneous wireless networks integration and broadband access. It is regarded as the potential core technology to build a next generation broadband mobile Internet. This thesis mainly concerns the joint routing and spectrum assignment technology of WLAN Mesh based on the cross-layer design idea under the circumstance of the Overlay spectrum sharing model. It is significant for building a high safety, large capacity, high speed, wide coverage Wireless Mesh Networks. The main work of this paper can be summarized as follows:Firstly, this thesis studies the cross-layer design model of routing and channel assignment in Cognitive Wireless Mesh Networks. The conventional technology of routing and channel assignment adopts an independent and bold design, so it is hard to update and meet special QoS demands. In this paper, SA cross-layer reference model is designed,combining with new characters in CogMesh: cognitive ability and reconfigurable ability. SA can intelligently evaluate and adjust the design of the current system with the cognitive reasoning and QoS identification ability.Secondly, this thesis studies the joint on-demand routing and channel assignment (QoS CCAR) in Cognitive Wireless Mesh Networks. With regards to the power interference and QoS guarantee problems of two kinds of users, INSIGNIA System is introduced to sense spectrum in the process of on-demand routing. When routing, the requirement information of application layer is collected. Channel assignment is done according to the routing metric from the delay-constrained link capacity. The simulation results indicate that the agreement can achieve better throughput and delay performance, support the flow restoration and DiffServ according to service styles.Finally, to cope up with the problems of node mobility and spectrum change in cognitive network, the joint table-driving routing and channel assignment (CR-OLSR) is designed based on OLSR agreement. This scheme makes full use of the topology information of IP layer, link quality and load information of link layer, and channel state information of physics layer to optimize the choices of MPR node and to finish the channel assignment in the routing process. Simulation results show that CR-OLSR can improve the packet delivery fraction at a little delay cost while the topology changes slowly.