Effectiveness of Multilayer Perceptron for Indoor Localization in Wi-Fi Enabled IoT Environments

Wi-Fi-based Indoor Localization is particularly important because Wi-Fi networks are ubiquitous and already present in many IoT indoor environments. Leveraging existing Wi-Fi infrastructure for localization purposes eliminates the need for additional hardware, making it a cost-effective and accessible solution. Multilayer Perceptron (MLP) plays a pivotal role in addressing the need for Indoor Localization in Wi-Fi-based IoT environments. Multilayer Perceptron (MLP), type of neural network, excels at capturing complicated relationships within data, making it well-suited for the dynamic and complex nature of indoor environments. Received Signal Strength Indicator (RSSI), Euclidian Distance(ED) measurements used for the distance estimation of devices connected to access points A, B, and C. The performance evaluation conducted through the metric of Mean Absolute Error (MAE), Mean Squared Error (MSE) and variance. Notably, the achieved optimal performance 0.8610 meters, underscoring the effectiveness of Multilayer Perceptron (MLP) in accurately estimating distances in diverse indoor environments. A significant Mean Absolute Error (MAE) improvement of 42.65% and 33.92%, respectively, is observed in Scenarios 1 and 2 when compared to the reported research work. There is a notable Mean Squared Error (MSE) enhancement of 25.04% and 30.15%, respectively, in Scenarios 1 and 2, compared to the reported research work.