Path loss models with distance-dependent weighted fitting and estimation of censored path loss data

Path loss models are the most fundamental part of wireless propagation channel models. Path loss is typically modelled as a (single-slope or multi-slope) power-law dependency on distance plus a log-normally distributed shadowing attenuation. Determination of the parameters of this model is usually done by fitting the model to results from measurements or ray tracing. The authors show that the typical least-square fitting to those data points is inherently biased to give the best fitting to the link distances that happen to have more evaluation points. A weighted fitting method is developed that emphasises the accuracy at the distance range that is consciously chosen by the user as most important for a system simulation. As a further important point that is typically not taken into account for path loss parameter extraction, the authors show that typically measurement data (but also ray tracing) is censored, i.e. path loss values above a certain threshold cannot be measured. The authors present examples of weighted fitting models, and models with and without the censored data, for 28 GHz channels in urban macrocells, and show that these effects have a significant impact on the extracted parameters and that the fitting accuracy can be improved with the presented methods.