Hybrid prediction model for field strength with ray tracing and artificial neural networks

In this paper, a hybrid scheme based on ray tracing method and the artificial neural networks for the field strength prediction is presented for the indoor environment. In the proposed scheme, ray tracing technique is used to coarsely predict the field strength according to the easily-acquired low accuracy geographical databases for the indoor scenarios. Then, the artificial neural networks are trained using measurement data or simulation data from fine geographical databases to compensate the effects caused by the detail geographical information in the propagation environments. In this way, high prediction accuracy is achieved with less running time. Simulation results reveal that using this hybrid method, the field strength is predicted in higher accuracy with less computation time.