Statistical Analysis and Modeling of User Micromobility for THz Cellular Communications

Terahertz (THz, 0.3-3 THz) wireless access is nowadays considered as a major enabling technology for sixth generation (6G) cellular systems. To compensate for extreme propagation losses these systems will utilize antenna arrays with extremely directional beams. The performance of such systems will thus be heavily affected by micromobility such as shakes and rotations even when user is in stationary position. The ultimate effect is spontaneous degradation of signal-to-noise (SNR) level leading to outages. In this paper, we measure and statistically characterize the micromobility process of various applications including video viewing, phone calling, virtual reality viewing and racing game. Particularly, we characterize occupancy distributions and first-passage time (FPT) to outage for various antenna configurations. We also assess the radial symmetry in micromobility patterns and characterize distance-dependent velocity and drift to the center parameters. The obtained results are essential for developing mathematical models of micromobility patterns that needs to be further used in system-level analysis of THz cellular systems. To this end, we also illustrate that Markov models are only suitable for applications with low and purely random micromobility dynamics such as video viewing and phone calling. When a user is controlled by the application, as in the case of gaming, Markov models overestimate FPT to outage.