A time-domain propagation model of the UWB indoor channel in the FCC-compliant band 3.6 - 6 GHz based on PN-sequence channel measurements

We derive a statistical model of the UWB indoor channel based on the experimental data collected in a modern office building. Measurements were made in different rooms throughout the floor and within each room the receiver antenna was moved over a square grid of 25 × 25 locations spaced 2 cm apart. The measurement technique was based on the use of a carrier at 4.78 GHz modulated by a train of short duration (0.4 ns) pulses shaped by a PN-sequence. Thus the probe signal covers the band 3.6 - 6 GHz. We coherently demodulate the received signals and cross-correlate their in-phase and q-phase components to an opportune PN-sequence template to extract the channel impulse responses from the recorded profiles. Then we post-process these "recovered" impulse responses by best-fit procedures to set up a statistical tapped delay line model of the UWB indoor channel. We model the path loss for LOS and NLOS conditions by distance power laws and the shadowing by log-normal distributions. The average power-delay profiles exhibit a clustered structure, which means that rays arrive at the receiver in groups, each having a given decay constant. We characterize the small-scale statistics by selecting the distribution that verifies with a 95%-confidence interval both the chi-square test and the Kolmogorov-Smirnov test applied to the experimental data. The gamma distribution verifies the abovementioned tests in most cases. The shape parameters of such gamma distributions spread in the range of 1 ÷ 3, and remain roughly constant around 2 with the excess delay.