A field strength prediction model based on artificial neural networks

A new field strength prediction model for mobile phone environment is presented. The model is based on the principles of very popular feedforward neural networks-multilayer perceptron. The implementation of this model relies on two databases: (1) a terrain elevation database and (1) a groundcover ("clutter") database. In order to build the model, extensive field strength measurements were carried out in a wide area of Belgrade, Yugoslavia, a typical European town of a size of 2 million inhabitants. CW signals of several actual transmitters operating in the 450 MHz and 900 MHz frequency bands were measured. Over 1,500,000 records of signal level versus position were collected. A comparison between data obtained by the field strength prediction, based on the proposed model and measurement data not taken into account during the development of this model, has shown that the proposed model is fast, accurate and reliable.