A wireless propagation channel model with meteorological quantities using neural networks

Deterministic channel modeling approaches are slow to run, require a detailed description of the environment (which is sometimes expensive or even impossible to obtain) and may be difficult to implement. A new approach for the modeling of wireless propagation in LOS environment is presented. We treat the meteorological conditions by weather variations through using neural networks. The aim of the paper is to propose a neural model for understanding the relation between the path loss, the propagation delay and the atmosphere parameters such as humidity, pressure, temperature. It is clarified the propagation factors affecting the wireless channel in the frequency range 300 MHz to 100 GHz. We use grey box approach based on fundamental principles of radio wave propagation physics and measurement data. To verify the accuracy of the model, evaluation and validation of the model are performed by simulating the channel using different sets of actual data from different situations. It is shown that this model can handle unusual atmosphere conditions and the model can be applied to better calculate the delay propagation.