A Dynamic Queue Management Algorithm for IEEE 802.11e EDCA Wireless LAN

IEEE802.11e enhanced distributed channel access (EDCA) standard classifies data traffic into different priorities to support varying QOS requirements, each access category (AC) has its own buffer queue and values of EDCA parameters. However, the method to manage AC buffer queues is static and not adaptive, when traffic of different AC are not balanced, it will result in unnecessary transmission delay and packet losses, when traffic is heavy, QOS requirements of some high priority services will not be meet. In this paper, we present a dynamic queue management scheme, the main idea is adding AC distributor in medium access control (MAC) layer, which can deliver various data traffic to different queues dynamically according to network load, transmission priority or buffer queue length. When a data frame from upper layer arrives to distributor, expected transmission delay of each AC is calculated according to previous transmission indicators, and then, the frame is delivered into appropriate AC queue, so packet loss rate and transmission delay can be reduced, network throughput can also be improved. The theory of our algorithm are introduced, performance indicators such as saturation throughput, transmission delay and channel access delay are analyzed by Markov random model. The analysis results appear that our scheme is an efficient way to improve the usage of wireless channels and throughputs with high QOS.