Research On Improving End-to-end Performance In Wireless Mesh Networks

WMN(Wireless Mesh Networks)is a new kind of broadband Wireless network and it has already begun into the form, including family net, enterprise net, public net and intelligent net applications. With the more applications of WMN, people need much higher network performance requirements. The major MAP select algorithm is based on RSSI, STA selects the MAP with the largest RSSI to access to network. But this MAP selection algorithm doesn't take into account the competition between the STAs, and didn't consider the current load of the MAP, some MAPs with smaller RSSI value are idle, so it can't effectively utilize the network resources. Because of the features of WMN, the MAP select algorithm should not only consider the quality of the link between STAs and the MAPs, but also need to consider the quality of path between the MAP and the gateway.This paper propose a MAP select algorithm which takes into account the competition between the equipment and the load conditions of MAP ,to improve stations'end-to-end communication performance with the Internet. Because WMN is dynamic, the optimal scheme introduces re-association mechanism in order to make it as the best MAP. In addition, introduce oscillation avoidance mechanism to prevent the occurrence of frequent re-association which caused performance degradation.We implemented the MAP selection algorithm in NS2 and ran a lot of simulations to validate our proposal. Simulation results show that the proposed joint association mechanism with the ENLAETT metric can significantly improve the end-to-end performance for wireless mesh networks by up to 60% than the traditional MAP selection algorithm based on the RSSI. And it is well adapted to changes of network conditions.