Characterization of Angular User Movement Based on Uplink Antenna Array Measurements

Sammanfattning

In massive MIMO and beamforming-based cellular networks, user mobility afects the channel through changes in the angular relationship between the user equipment (UE) and the base station. This thesis uses live uplink antenna array data to study user mobility in the angular domain. This data consists of Sounding Reference Signal (SRS) channel measurements from operational cellular sites. The channel data was divided into overlapping snapshots and converted into covariance matrices. Direction-of-arrival estimates were obtained by evaluating beamforming power over candidate steering vectors. An FFT-based method was used for the large-scale analysis, and the resulting spatial direction trajectories were fltered using a Kalman flter before angular velocities were computed. The results show that the observed angular mobility is dominated by low velocities, with most samples corresponding to little or no angular motion. The horizontal angular velocity is consistently larger and more variable than the vertical component. The spatial analysis further shows that mobility is not uniformly distributed across the antenna feld of view. Higher angular velocities occur mainly in the lower part of the angular grid and in cells containing active infrastructure such as roads or railways. A simple mobility simulation based on the Gauss-Markov mobility model reproduces broad trends such as low overall mobility, stronger horizontal motion and non-uniform spatial distributions of motion, but it does not capture localized site-specifc structures. These results show that angular mobility extracted from uplink array measurements contains meaningful spatial and cell-dependent structure. They also indicate that realistic simulation of beamforming and channel aging should account not only for user speed distributions, but also for how the physical environment shapes angular motion as observed from the base station.