Research On Routing Protocol With Multiple QoS Guarantee Based On Ant Colony In Wireless Mesh Network

Wireless Mesh Network (WMN) is a wireless network with multi-hop, high-speed andhigh-capacity. It is the "last mile" broadband access solutions which combines the characteristics ofthe WLAN and Ad hoc networks. As one of the core theoretical research issues of the next generationwireless, it is a challenging proposition to provide transmission services with QoS guarantees. Withthe development of network, the popularity of multimedia services, as well as the needs ofcommercial applications, the demanding of WMN’s providing QoS guarantees is increasing, so it isnecessary to focus on a further research on the QoS routing of WMN.It is very difficult and challenging to provide QoS guarantees for multimedia service in WMNbecause of some features of WMN itself such as complexity and uncertainty of Wireless channel andlimitation of mobile terminal. Implementing QoS routing in the WMN has been proved to be anNP-Complete problem. Considering that heuristic algorithm has a good effect on solvingNP-Complete problems, a WMN routing algorithm based on ant colony is put forward in this paper.The concept of Wireless Mesh Network and ant colony algorithm are introduced firstly in thispaper. Then the difficulties in implementing QoS routing are analyzed. The following is someintroduction about the research status of routing protocol of single QoS guarantee and multi-QoSconstraints. Considering that implementing QoS routing is an NP-Complete problem, a PH_AODVrouting algorithm with load balance is proposed. PH_AODV converts multiple QoS constrains to auniform dimension which is integrated into the pheromone update strategy according to therequirement of transmission services. And in the routing process, the load of neighbor nodes isconsidered as the path selection, which makes the protocol have congestion control function. In thesimulation experiment, it is demonstrated that this protocol has a better performance than AODV andAOMDV. In addition, given the lack of corporation closeness of ant colony, a kind of diffusion modelwith pheromone based on PH_AODV protocol is proposed in order to enhance cooperation among theants. Simulation experiments demonstrated that pheromone diffusion model can improve networkperformance efficiently.