A 3D Geometry-Based THz Channel Model for 6G Ultra Massive MIMO Systems

In this paper, a 3D geometry-based double-spherical terahertz (THz) channel model is proposed for ultra massive multiple-input multiple-output (MIMO) communications. To reflect real THz ultra massive MIMO communications, the characteristic of nano-material array antenna and the high loss of THz bands are considered in channel modeling. Based on the proposed geometry-based stochastic model (GBSM), power delay profile (PDP), space-time-frequency correlation function, and the Doppler power spectral density are derived and analyzed. Furthermore, the proposed model is validated by comparing the simulated results with the measurement reference. Influences of channel parameters, namely, antenna element spacing, carrier frequency, number and location of clusters, on the correlation functions are investigated. The results indicate that antenna element spacing and number of clusters have significant impacts on channel correlation, and the decreasing of correlation function becomes slower when antenna element spacing decreases. Moreover, location of clusters also affects the drop rate of correlation function. The influences of the number of clusters and concentration of intra-cluster multipath components (MPCs) on the Doppler power spectral density are analyzed. It is found that a larger number of clusters and a more concentrated distribution of intra-cluster MPCs increase the degree of Doppler spectrum variation. The proposed channel model and the corresponding statistical properties are insightful for designing and realizing THz ultra massive MIMO systems for 6G and beyond.

Automated summary

This paper proposes a 3D geometry-based double-spherical terahertz (THz) channel model for ultra massive multiple-input multiple-output (MIMO) communications. The model takes into account the characteristics of nano-material array antenna and the high loss of THz bands. The influences of channel parameters on correlation functions are investigated and it is found that antenna element spacing and number of clusters significantly impact channel correlation. Moreover, the location of clusters affects the drop rate of correlation function. The influences of the number of clusters and concentration of intra-cluster multipath components (MPCs) on the Doppler power spectral density are also analyzed.