I’m running simulations varying both the tilt and the height of the UE in order to see how the throughput is affected. However, the variation of the throughput is minimum, and it doesn’t hold a tendency. Given the fact that the pathloss model depends on the UE height, I thought that the throughput would be affected and that it’d be noticeable, but that doesn’t seem to be the case. Could you please explain to me why this happens? Also, on what does the throughput variation depend?
On avergage you will see no improvement on the throughput when you vary both the user height and the electrical dowtilt. This comes from the fact that the electrical downtilt value is not the appropriate one for different user heights.
If you analyse a single user, you will see a difference in throughput, depending on the electrical downtilt value, distance from BS station, user height, e.g. choosing an electrical dowtilt value that corresponds to the user height, the beam pointing in the direction of the user, you should see the highest throughput. In the above mentioned paper, such simulations are conducted and results in terms of throughput are shown.
1. The electrical downtilt is the value of electrical downtilt per antenna array element, that is used to form the beamforming weights in the analog domain -steer the beam. Have a look in +channel_gain_wrappers/TR36873_Fading_3D_channel.m: function calculate_channel_coefficient_NLOS, how the parameter element_weight is defined.
2. Slant angle denotes the mechanical rotation of each antenna element around z-axis. This value is specified in 3GPP TR36.873 Table 8.2-2. to be zero at the user side, while at the BS antenna you can modify it depending on how you consider the mechanical position of an antenna element. If the slant angle is set to zero, the antenna element lies the z-axis, i.e. a vertical dipole.
If you still see no gain/difference in throughput for different parameters, you might be in saturation, low or no interference, large SINR.
Let me know if you still have troubles in this issue.