Measurement and Modeling of Angular Spreads of Three-Dimensional Urban Street Radio Channels

By utilizing planar antenna arrays, a base station (BS) by the roadside can exploit two-dimensional (2-D) sectorization and beamforming to connect dense pedestrians and vehicles in a street. The azimuth and elevation angular spreads of the street radio channels are critical for the spatial multiplexing performance. In this paper, by utilizing a multiple-input-multiple-output (MIMO) channel sounder equipped with two uniformed planar antenna arrays, the three-dimensional (3-D) multipath propagations in urban macrocellular street canyon environments were measured and modeled, with the focus on the dynamics of the angular spreads along the streets. The transmitter (emulating a user equipment, UE) was placed at 100 and 95 positions in two streets for the line-of-sight (LOS) and non-line-of-sight (NLOS) scenarios, respectively. The azimuth/elevation angle-of-arrival (AoA/EoA) and the root-mean-square azimuth/elevation spread of arrivals (ASA/ESA) at each position were measured. Contrary to the expected monotonous change of the angular spreads with respect to the UE-BS distance, the measurement results show that, in the LOS scenario, ASA and ESA have a positive correlation with the UE-BS distance in an open street but a negative correlation in a closed street. In the NLOS scenario, the correlation is positive when the UE is close to a building without the over-rooftop diffraction, but the correlation is reversed if such a Quasi-LOS path exists. The 2-D arrival profiles of the ray clusters have been observed, and their impacts on the angular spreads are analyzed in different propagation environments. By comparing multiple candidate fitting functions, the lognormal distribution models for ASA and ESA are proposed. In addition, the channel delay spread (DS) was also measured along the streets and positive correlations among ASA, ESA, and DS have been found. This work can help to establish the 3-D spatial channel models for advanced MIMO technologies and is also valuable for future channel measurements.