Abstract

When performing wireless network simulations, the lack of precise channel modeling in simulator frameworks becomes a serious problem. Often deterministic models are used for packet propagation, which describe real conditions insufficiently. To close this gap we extended the OMNeT++ Mobility Framework to support probabilistic propagation models. We provide an implementation for the Log-Normal-Shadowing, Nakagami, Rayleigh and Rice wave propagation models and set up a framework that allows easy integration of additional models in future. Due to the characteristics of probabilistic radio models a fixed maximum packet propagation range encounters the problem of inaccurate simulation results as relevant events may be suppressed. On the other hand, unlimited packet propagation, which guarantees for correct simulation runs, causes unnecessary simulation overhead. In this work we present an approach to limit the event delivery to the area where the probability that the event is relevant to the simulation exceeds an adjustable threshold. In order to validate our extensions we successfully performed a detailed crosscheck with the network simulator NS-2 and run a performance evaluation and comparison.