Wlan Performance Inprovement For Dense Deployment Scenarios

Along with the booming development of Mobile Internet, the number of WLAN devices has experienced an explosive growth. For the consideration of network coverage and transmission capacity, a large number of APs have been deployed in our streets, offices and residential areas. Each of APs acts as a Basic Service Set (BSS), and provides network access for our mobile stations such as laptops and smartphones. Some of those APs are Fit AP (also called Thin AP) managed by an Aceess Controller (AC), the rest of them which have full functionality and can managed by themselves are called Fat AP. Due to the lack of unified frequency planning and transmit power control, it’s hard to prevent some of those APs from using the same channel when they are deployed by large scale. When the APs which use the same channel have to overlap in their coverage, thereby the Overlapping BSS (OBSS) problem arises. Under the condition of limited frequency resources and the lack of effective transmit power control scheme, the dense AP deployment may bring bad effects such as more complicated interference situation or lower quality of experience (QoE) instead of performance inprovement.In this thesis, firstly, existing WLAN managemnet architecture is discussed, from which, we could reach a point that the future WLAN would develop into "Terminal-Cloud Fusion" architecture. Then the Dense WLAN is put forward, which is then divided into "Dense STA Access" and "Dense AP Deployment" two aspects. The detailed interference analysis is provided by system level simulation. Then possible improved solutions are discussed based on the analysis. Considering that Fat APs as a large proportion of APs used in actual scenario without effective management strategies, this thesis would mainly focused on the performance improvement of dense WLAN with Fat APs. Driven by this purpose, a Distributed Dynamic Channel Allocation (DDCA) algorithm is proposed based on an actual research project. The algorithm mentioned in this thesis has been realized in Linux platform, which is based on a concept named "Wireless Channel Duty Ratio". Finally, simulation and actual test are carried out to evaluate the performance improvement of this algorithm.