Planning a radio network (in any technology: LTE, WiMAX, UMTS, GSM, TETRA etc.)
requires knowledge of where and how to place the sites in order to provide the radio coverage which will provide the service. In order to estimate what the radio coverage will be a PREDICTION is made using a propagation model. A propagation model is a technique (usually a mathematical equation) which brings together all (or as many as one would like to consider) the parameters which would (or could) affect the radio signal – frequency, antenna height, receiver height, distance, path profile, angle of incidence, etc. There are many “standard” models - Cost231, SUI, Okumura-Hata etc, which try and predict propagation under certain conditions and within certain limits dictated usually by the parameters and conditions under which the models were derived. Another option is to derive a propagation model for a particular scenario or condition – and this is what a model calibration exercise tries to achieve. The advantage of this is that the model which is derived should be accurate and therefore the predictions of coverage will be accurate.
Now the model (along with the underlying accuracy of associated data such mapping data, accuracy of source CW data etc) is the cornerstone in any further analysis of the network such as site count, coverage and service quality levels, so a good and accurate model will enable a much more accurate prediction the behavior of the network and the quality and cost.