Investigation of different fading forecast schemes for flat fading radio channels

After a brief introduction explaining the motivation for predicting the mobile radio channel both a deterministic channel model and an autoregressive signal model as well as corresponding parameter estimation schemes are introduced, which can be used for a fading forecast of flat fading radio channels. A preliminary performance test using Jakes´ model is carried out for having a fundamental idea of the performance and the properties of these two models and the different estimation schemes. For more representative results, the most appropriate methods are investigated using data of a huge amount of different radio channels generated by simulations using the statistical model of Dersch. In order to verify the simulation results, the algorithms are further applied to measurement data of a time-variant indoor radio channel. With realistic simulation data or measurements, respectively, the autoregressive model performs best. Depending on the chosen parameter estimation scheme a prediction range of more than one wavelength ahead is feasible. Furthermore, the simulations reveal the common applicability of the proposed prediction schemes in different applications of wireless communications.