Study On Distributed Power Control In Wlan Based On Ieee 802.11 Protocols

In this thesis we mainly study on the power control in WLAN based on the IEEE 802.11 protocols. Power control here is a concept of Radio Resource Management. With a background project which studied on the scene such as 2010 World Expo Shanghai,large-scale expos have some characteristics such as asymmetry on the time and space scale. Because the environment changes fast and the place to cover is so wide that it needs distributed schemes and algorithms to accomplish the timely optimization.In the beginning we introduced the power control in WLAN, including the optimization targets, contents and some existed schemes and algorithms. Then we analyzed the factors that will worsen the performance of a IEEE 802.11-based WLAN. We also classified the power control method on each OSI layer and focused our work on the MAC layer.We proposed a completed distributed scheme to accomplish the optimization in such WLAN, aiming at reducing the interference among APs and boosting the whole throughput. We discussed almost the every detail in each flow in the scheme and the relevant problems which the design aims at, then we proved the correctness and validity by simulation on boosting the whole throughput and reacting the fast change of environment.Then we discussed some more general situations of two important parameters in our scheme: the linkage gain of wireless channel and the saturated throughput. Then the prediction of gains in multipath attenuation channels is discussed, using some mature models and math methods like autoregressive prediction. For saturated throughput we introduced two difference existed way to analyze the 802.11 DCF mode access mechanism and the relevant result. Based on the result we proposed a reward-cost function with the throughput and power consume. We could solve from the function to obtain the optimized physical rate of each nodes in the network.After that we proposed a schedule algorithm for solving the leak on wireless covering, as a potential outcome of the power control. We concerned the fact that the traffic on a single source is always short range dependent and transmitted in burst mode. So we used the heavy-tailed distribution to model the source traffic arrival, and adopted Markov Modulated Bernoulli Process model to predict the probability of the new arrival request out of the covered area. Then two methods were proposed to make decision on how to enlarge the area to capture the potential request. The simulation had showed the performance on the capture ratio.In the end of the thesis we concluded the work and gave a view on the future study.