A Learning Based Rate Adaption Algorithm in 802.11n Networks

Rate adaptation algorithms play an important role in improving the performance of WLANs. However, just few numbers of efforts have been made on this area. These algorithms try to provide more throughputs by increasing or decreasing rates upon channel condition and obtained good put However, it is a little different in 802.11n. Unfortunately, just a few efforts have been made on this issue. This paper presents a learning automata based algorithm, called learning_RA, to determine the best rate in 802.11n networks. As it is shown in the results, our proposed algorithm provides better results and it has a simple implementation too. We provided a comparison of three different states in the same scenario, that is, without rate adaption, using zigzag and learning_RA algorithms. Obtained results imply that the proposed algorithm outperforms zigzag, which is a well-known rate adaption algorithm.