A measurement data-based investigation of fading modeling for indoor THz wireless systems

Abstract

The terahertz (THz) band offers a vast amount of still unallocated bandwidth, which makes it a promising enabler for future sixth generation wireless systems. The high frequencies of the THz band lead to a significantly reduced multipath richness of the propagating THz signals. However, there are still paths that can carry a significant amount of power. As a result, the THz band small-scale fading characterization is of particular interest and the appropriate stochastic distributions that best fit the empirical channels need to be identified. This work investigates the suitability of the 'alpha-mu', Rice and Nakagami-m distributions to adequately model the small-scale fading statistics of the channel gain measurements of various Line-of-Sight (LoS) and Non-Line-of-Sight (NLoS) indoor THz wireless links. The fitting accuracy of the examined analytical distributions is validated by means of the Kolmogorov-Smirnov test, the Kullback-Leibler divergence, the logarithmic Kolmogorov-Smirnov test and the root-mean-square-error. Also, the ergodic capacity based on the channel gain measurements as well as on the 'alpha-mu', Rice and Nakagami-m distributions is presented. Based on the fitting accuracy metrics the Rice and 'alpha-mu' yield the best fit for the LoS and NLoS links, respectively. The Nakagami-m does not fit the empirical distributions for any of the presented links. Furthermore, insights are provided for the ranges of the extracted values of the analytical distributions in LoS and NLoS transmission conditions.