The Study Of Centralized AP Selection Mechanism And Power-Saving Scheduling In WLAN

Wireless Local Area Network(WLAN) has been developed over 20 years and becomes a widely applied technology due to it's mat uration, stability, easy expansion and convenient management. As the standard of WLAN, the IEEE 802.11 protocol family has been one of the most used network protocol.The conventional approaches in IEEE 802.11 typically use Received Signal Strength Indicator(RSSI) as the AP selection criterion without considering the load of each AP. This type of scheme results in a significant load imbalance between several APs and causes the decreasing of network performance and resource utilization rate in the network containing numbers of APs. To solve this problem, we proposed an centralized AP selection method based on Software Defined Network(SDN), which has the ability to improve network resource utilization, increase t hroughput and save the Stations'(STA) electricity. Firstly, the function can grasp the information of the entire network with the help of controllability of SDN, which makes the following choices more accurate. Secondly, STAs cost little electricity in the function to select APs, for that the task of information processing and calculating is mainly on the SDN controller. Thirdly, the proposed selection mechanism is ordered by APs' load level and STAs' activity and therefore more efficient. We experiment a simulation and get the result that the function could highlight throughputs and decrease switching frequency with more accurate selection, compared to other proposed methods.In a large-scale WLAN, sometimes lots of STAs need to upload or download fixed data, such as the meter reading service in the smart grid. Due to the huge number of STAs, the Distributed Coordination Function(DCF) will lead to worst transmission efficiency. To accommodate the application characterized by simultaneous service, short data frames, the same format and frequent transmission, we propose a power-saving scheduling method. The method is based on the Power Save Multi-Poll(PSMP) mechanism and schedules all the STAs to receive and transmit data, which avoids extra overheads from channel competition. At the same time, the method can greatly compress the information of STAs, and reduce the overhead of scheduling, compared to the PSMP protocol. So, this method is absolutely fit to the WLAN containing great number of STAs and able to achieve higher scheduling efficiency and lower power consumption. In this thesis, the frame structure and the scheduling process are described in details. The simulation evaluates the performance of our proposed method in terms of overhead, activation time and overall throughput. Obviously, the result shows that our method meets the power-saving requirements of the large-scale WLAN.