Model Selection for Path Loss Prediction in Wireless Networks

Abstract

Path loss prediction is an important task in mobile communication networks. Quality of communication between nodes depend on the environment in which the network is operating. Path loss occurs due to many effects such as free-space loss, diffraction, refraction, and reflection. In this paper, we apply different machine learning techniques to model the path loss and to predict the loss in a similar environment. We have used distance vs signal strength data from different wireless access points. The comparison of different models shows that Kalman filtering is performing better in predicting the path loss.