For a long time researchers have been using the stationarity assumption in order to make use of WSSUS channel models. In the last decades such channel models were successfully used to simulate for example the GSM channels.
Recent measurements for mobile-to-mobile channels (car2car, air2air) have revealed that channels with two moving entities don’t exhibit stationary behaviour. Especially for head-on scenarios the channel is highly time-variant. Therefore purely stochastic channel models should not be used in this case.
At the moment a lot of research is done in the field of mobile-to-mobile communication systems. For this reason it is necessary to have accurate channel models for those types of propagation scenarios. Especially intelligent transportation systems in the vehicular sector, which are used for traffic flow optimization or collision avoidance, will play an important role in the future. 
In this talk we explain how a novel geometric-stochastic channel modeling technique allows us to incorporate the non-stationarity into our channel model. We analytically derive time-variant, delay-dependent Doppler probability density functions for both vehicle-to-vehicle and aircraft-to-aircraft channels. Hereby the aircraft-to-aircraft channel can be seen as an extension of the vehicle-to-vehicle channel.
To validate our theoretical model, we also conducted two measurement campaigns. One of these campaigns will be presented in detail and the results will show a good agreement with the theoretical model. 
With the time-variant pdfs the mobile-to-mobile channels are modelled more accurately and the non-stationarity is naturally included as well. This will make mobile-to-mobile channel simulations more realistic, so that the system’s performance estimate becomes better.