In radio network performance evaluations it is of interest to use simulation environments that represent realistic networks. In this paper two widely used simulation environments are benchmarked against live network path gain measurements. Some parameter adjustments are done to fit the simulation environments to the measured path gain and cell isolation statistics. The need to modify base station antenna model parameters to match the intra-site cell isolation in real networks is identified and elaborated. In particular an increase of maximum attenuation to 35dB, introduction of mechanical tilt and reduction of vertical beam width are needed. In addition some adjustments are done to better represent the assumed traffic distribution with 80% indoor users and expected propagation losses resulting in a path gain level on par with the measured network.
Arne Simonsson, Martin Johansson and Magnus Lundevall