A Study Of Association Control Load Balancing In 802.11 Wireless Lans

Recently, wireless local area network (WLAN) has developed rapidly with the superiority of mobility, flexibility and expansibility. As the application of WLAN extends unceasingly, WLAN is facing with more and more grave challenges in realistic deployment. In 802.11 standard, the selection policy of Station (STA) is merely based on the receive signal strength indication (RSSI), which makes STAs select the Access Point (AP) with strong RSSI intensively. It causes the loads among APs distribute unevenly, leading to bad resource utilization. Besides, 802.11 protocols support multiple data transfer rate. The STA with lower rate would affect those with higher rate and then degrade the global throughput. Therefore, how to balance the loads among APs and maximize the network performance has become an urgent problem to be solved. Meanwhile, along with the continued expansion of the scale of WLANs, the disadvantages of the traditional Fat AP architecture has become more and more obvious on the development of large-scale WLANs. So the new Access Controller (AC) + Fit AP architecture has appeared on the scene. Confronted with the new network architecture, how to unified control the loads of network, and achieve the balance of the loads in global, has become one of the research hotspots.This thesis focused on the multi-rate characteristic of 802.11 WLANs and analyzed the Distributed coordination function of MAC layer expansively. Then, we took cognizance of the phenomenon that the STA with lower rate would affect those with higher rate and then degrades the global throughput, and proposed a Distributed Adaptive Load Balancing Algorithm for Multi-Rate WLANs (DALBA), which took into account of RSSI, throughput and data transfer rate. Beside, this thesis focused on the new FIT AP architecture, analyzed the Control and Provisioning of Wireless Access Points (CAPWAP) protocol and studied on the OpenCAPWAP open source project. Then, we extended the CAPWAP protocol according to our need, and proposed a Centralize Control Load Balancing Algorithm Based on CAPWAP (CCLBA). Both algorithms adopted the association control load balancing mechanism. Additionally, we examined our algorithms with Opnet Modeler 14.5, and compared with the Best RSSI First (BRF) policy of the 802.11 standard. The result showed that both of our two algorithms could upgrade the network performance and perform better than the BRF policy.