Abstract
This paper proposes a low-complexity hybrid beamforming design for multi-antenna communication systems. The hybrid beamformer comprises of a baseband digital beamformer and a constant modulus analog beamformer in radio frequency (RF) part of the system. As in Singular-Value-Decomposition (SVD) based beamforming, hybrid beamforming design aims to generate parallel data streams in multi-antenna systems, however, due to the constant modulus constraint of the analog beamformer, the problem cannot be solved, similarly. To address this problem, mathematical expressions of the parallel data streams are derived in this paper and desired and interfering signals are specified per stream. The analog beamformers are designed by maximizing the power of desired signal while minimizing the sum-power of interfering signals. Finally, digital beamformers are derived through defining the equivalent channel observed by the transmitter/receiver. Regardless of the number of the antennas or type of channel, the proposed approach can be applied to wide range of MIMO systems with hybrid structure wherein the number of the antennas is more than the number of the RF chains. In particular, the proposed algorithm is verified for sparse channels that emulate mm-wave transmission as well as rich scattering environments. In order to validate the optimality, the results are compared with those of the state-of-the-art and it is demonstrated that the performance of the proposed method outperforms state-of-the-art techniques, regardless of type of the channel and/or system configuration.