Client-Network Collaborative Load Balancing Mechanism for WLAN Based on SDN and 802.11U

Xiaokang Sang,Qian Wu,Hewu Li
DOI: https://doi.org/10.1109/iwcmc.2017.7986337
2017-01-01
Abstract:As the rapid growth of Wireless Local Area Network (WLAN), the densities of both Access Points (APs) and clients have dramatically increased. In the dense and crowded WLANs, the challenges of load balance become non-negligible due to the problems raised by traditional AP selection mechanism based on signal strength. To address this issue, many existing works consider the load metrics and conduct dynamic topology adjustment for the AP selection. In these literatures, a specific server and trick communications between client and network are always introduced, but they are not general. In this paper, we propose a novel load balancing mechanism, which is implemented on a client-network collaborative architecture based on Software Defined Network (SDN) and 802.11u. In our architecture, we implement several SDN applications to collect and maintain the information of load and topology, and design Network Resource Query Protocol (NRQP) to exchange the metrics of AP selection between client and network based on 802.11u. With the help of this architecture, we propose a novel load aware algorithm which can take the comprehensive consideration of the link state, client traffic and AP load. It can not only work when first association or moving handoff, but also conduct topology adjustment at proper time. We evaluate our mechanism on a real SDN prototype testbed and the results show the competitive performance.
What problem does this paper attempt to address?