AbstractIn this paper, we study the optimization of two tilt angles corresponding to two antenna arrays in each base station (BS) of a massive multiple-input multiple-output system. We consider two scenarios with perfect channel state information (CSI) and imperfect CSI. In the limit of the number of the BS antennas, the channel orthogonality is employed to derive the limit and the lower bound of the throughputs. By maximizing the lower bound or the limit throughput, the two antenna tilt angles are optimized. Simulation results show that the throughput performance can be improved with the designed tilt angles.
A Safe Reinforcement Learning Architecture for Antenna Tilt Optimisation
