Parametric modeling of wideband and ultrawideband channels in frequency domain

This paper investigates parametric modeling of wideband and ultrawideband (UWB) wireless channels in the frequency domain. By uniform sampling of the channel frequency response (CFR), an equivalent discrete channel can be obtained. The channel impulse response (CIR) of the discretized channel can be obtained by applying inverse discrete Fourier transform to the uniformly spaced CFR samples. Both uniformly spaced CFR and CIR samples can be extended to periodic sequences with the same period. It is shown that parametric modeling in the frequency domain is possible, if and only if the uniformly spaced CFR samples satisfy the wide-sense stationary (WSS) condition that is equivalent to the uncorrelated scattering (US) condition for the discrete CIR. This new WSS-US condition is fulfilled if the underlying continuous-time wideband or UWB channel is WSSUS with independent path gains and arrival times. It is also shown that there exists an analytic relation between the power spectral density of the uniformly spaced CFR samples and the power profile of the discrete CIR that satisfies the WSS-US condition. The main goal of this paper is to demonstrate that the wideband and UWB channels can be adequately modeled by low-order autoregressive (AR) and/or AR moving average models in the frequency domain. The results on channel modeling are illustrated by simulation examples.