7.5.1 FD Modeling

In this case we shall consider two types of modulation, QPSK and 16-QAM. The system behavior shall be analyzed both in Rayleigh as well as Rician fading environments. We first need to cover certain basic concepts as regards the modus operandi of the simulations. Broadband mobile radio channels can be characterized as linear time-variant filters with a complex equivalent lowpass IR as h( ?,t) ( ? = excess delay; t = time variability). In the FD we use its Fourier transform H( f; t.). However, over a short time interval, the shadowing can be considered constant, and in such a case, we can characterize the channel as a WSSUS. This means that we can apply the theory of linear time-variant filters [25, 26]. This assumption is appropriate for describing the small-scale behavior of the channel, which is suitable and sufficient for our purposes. The nature of the channel will be frequency selective due to the high rates of transmission considered.

We can model the time variability of a narrowband flat fading signal by shaping the spectrum of a complex Gaussian noise process. This is called Jakes's fading model [27]. The resulting colored, complex...