W-LS-IR Algorithm for Hybrid Precoding in Wideband Millimeter Wave MIMO Systems
Fulai Liu
Ruiyan Du
Xiaodong Kan
Xinwei Wang
Hybrid analog/digital precoding is a promising technology that reduces the hardware complexity and power consumption of large-scale millimeter wave (mmWave) multiple-input multipleoutput (MIMO) communication systems. Most prior work has focused on hybrid precoding for narrowband mmWave systems. MmWave systems, however, will likely act on wideband channels with frequency selectivity. Therefore, this paper presents an effective OFDM-based hybrid precoding algorithm (named as W-LS-IR algorithm) for wideband mmWave systems. Firstly, the initial phases of the analog precoding matrix are randomly generated, and the digital precoding matrix is initialized via the least squares (LS) method. Then, the column of the analog precoding matrix is derived from the dominant left singular vector of a residual matrix, and the corresponding row of the digital precoding matrix is updated using the LS method. Through the iterations of the aforementioned stage, the hybrid precoding matrix will approach a stable solution finally. Compared with related works, the proposed algorithm can improve the spectral efficiency of wideband mmWave MIMO communication systems. Simulation results are presented to confirm the efficiency of the proposed algorithm.