A truncation method for modelling effective antenna patterns in urban areas

Signal level predictions for sector antennas in mobile networks strongly depend on the antenna pattern used in the prediction tool. For built-up areas, reflections and scanering often lead to a significant energy component at the back of the antenna, and the theoretical antenna pattern does not account for this effect. We propose a method that \´\´truncates" the modelled antenna pattern at the hack of the antenna, such that the Front-To-Back ratio of the pattern becomes smaller. A comparison of the method with memuremen6 shows that this leads to an improved accuracy of the signal level predictions in an urban area.