Performance Analysis And Improved Fairness Of DCF In Wireless Local Area Networks

Along with the advance of wireless technology, IEEE802.11 Wireless local area network(WLAN) has got a rapid development and gained widespread application by virtue of its strong ability to organize networks, mobility and extensibility. WLAN only relates to the physical layer(PHY) and media access control(MAC) layer of OSI model(Open System interconnection). Besides, the property of MAC layer is one important factor affecting the performance of WLAN. Since, DCF(Distributed Coordination Function) mechanism is the main way to implement MAC protocol. Therefore, it has a great meaning and prospect to focus on the performance research and analysis of DCF protocol.Based on a total familiarity with DCF protocol at MAC layer, the paper analyzed and compared a variety of common performance analysis methods of DCF protocol, including the analysis of the simulation experiment and mathematical model. Then we mainly studied the two-dimensional Markov chain model proposed by Bianchi. Through the observation of the characteristics of the model and the result compared with the simulation experiment, the deficiencies of Bianchi model were obtained. Then considering the actual network condition, an improved two dimensional Markov chain model was given, making it more accord with standard of IEEE802.11. The formula of saturation throughput and average access delay in the system are deduced, the effects on the IEEE 802.11 performance by different terminal number and transmit load are analyzed. The correctness of the improved model is verified by simulation experiments. And through the analysis result of access time delay, we got network performances of different channel rate in the mode of basic access and RTS/CTS mode, which offers a good reference on the access mode-selection under varied channel rates.The channel unfairness of DCF protocol was also studied. In IEEE802.11 wireless network infrastructure, AP(Access Point) and each mobile terminal has the same access probability, however, in actual network, AP may send data to multiple mobile terminals simultaneously while seldom mobile terminals upload data to AP, which creates serious unfairness of bandwidth allocation. The paper proposed a bandwidth allocation strategy between downlink and uplink based on the improved method of transmission with probability and achieved by adjusting the transmission probability of AP. We deduced the relation between the optimal probability and number of downlink and uplink, making sure AP has a higher priority in the channel, so as to ensure the fair distribution of resources. Both the theory analysis and practical result showed that this mechanism can greatly improve the resources unfairness of downlink and uplink, and can improve the network throughput on the whole.