NEURAL-NETWORK APPROACH TO PREDICTION OF TERRESTRIAL WAVE-PROPAGATIONFOR MOBILE RADIO

Theoretical investigations into the suitability of neural network simu lators for the prediction of field strength based on topographical and morphographical data are presented. Effective input and output data p rocessing is developed using deterministic and heuristic formulae for the training of a neural network simulator. First successful steps tow ards a planning tool for mobile radio communication based on adaptive learning from measurement data are made. Field strength prediction wit h the new system shows acceptable conformity with measurements.