Adaptation using neural network in frequency selective MIMO-OFDM systems

In this paper, we proposed a neural network (NN) framework as a machine learning technique for link adaptation based on adaptive modulation and coding in 802.11n MIMO-OFDM wireless system to predict the best modulation and coding scheme (MCS) index under packet error rate (PER) constraints. Our approach is compared with the k-nearest neighbour (k-NN) algorithm in frequency selective wireless channels. Simulation results validate the implementation of proposed neural network framework in frequency selective channels, and show that the neural network technique outperforms k-NN algorithm especially in terms of PER when low MCS index selection which provide higher communication reliability is exploited.