Performance of MLP NN Using CSI and Different Standard IEEE 802.11 with Variable Channel Width

The combination of Channel State Information (CSI) and neural networks has gained attention due to its ability to enhance accuracy and reliability in tasks such as WiFi localization, gesture recognition, and indoor positioning. This paper explores the impact of 20 MHz and 40 MHz channel widths and IEEE 802.11/n/ax standards on WiFi localization using Multi-Layer Perceptron (MLP) neural networks and CSI. To evaluate the performance, experiments were conducted in a static indoor environment. For collecting CSI data, the FeitCSI tool was used. The study investigates the impact of different factors, such as the number of hidden layers, number of subcarriers, and input features, on the performance of the MLP neural network.