Research On Multi-gateway Load Balance In Wireless Mesh Network

Wireless Mesh Network is a broadband wireless network, which can also be seen as acombination of WLAN and MANET. Network performance of WMN may decline with the increasingcomplexity of network topology and the number of nodes. WSN load imbalance is mainly reflected ininternal network nodes and gateway nodes. The performance of the entire network will be greatlyinfluenced by Load imbalance.Firstly, an overview of WMN which contanins related technologies of the network internal nodesload balancing and gateway nodes load balancing is provided in this paper. Then the key problems ofload detection method and typical multi-gateway routing protocols are introduced.Secondly, a congestion-aware routing protocol named CA_HWMP(Congestion-aware HWMP) isproposed in this paper. The calculation method of Node load is improved and the node load is dividedinto three levels. To make nodes know load of surrounding neighbors, Load information is added to802.11s management frames. Route discovery and route maintenance are improved through LoadConstraint Control Mechanism. Different updating methods are adopted for nodes in different loadstate when updating PREQ messages. It can balance network traffic and avoid congestion throughconsidering both routing metric and load of neighbors.Thirdly, a hybrid routing protocol of multi-gateway load balanced namedMCA_HWMP(Multiple Gateway Congestion-aware HWMP) is proposed in this paper on the basis ofCA_HWMP. A multi-gateway selection strategy through taking into account the remainingtransmission capacity and the transmission cost of the current mesh node to gateway nodes is putforward by MCA_HWMP. Data is distributed to multiple gateways to further enhance thetransmission capacity of the entire network. Thus, load balance of gateway nodes is achieved from theinside out.Finally, CA_HWMP and MCA_HWMP are simulated in NS2network simulator. The resultsshow that the improved protocols have excellent performance in prompting network throughputs,reducing end-to-end delay and packet loss rate, etc.