A Sub-band Precoding Scheme for Wideband Massive MIMO-OFDM Systems

Sub-band precoding (SBP) stands out as a promising technology for orthogonal frequency division multiplexing (OFDM)-based wireless communication systems where the frequency selective channels exhibit gradual fading characteristics. Under such environment, the complexity of coefficient design can be significantly reduced by segmenting subcarriers into smaller sub-bands and applying identical precoding coefficients to all subcarriers within each sub-band. In this paper, a SBP algorithm is proposed for wideband millimeter wave (mmWave) and terahertz (THz) massive multiple-input multiple-output OFDM (M-MIMO-OFDM) systems. By leveraging fractional programming, quadratic transformation and first-order Taylor approximation, we formulate sum-rate maximization as a convex optimization problem, enabling efficient derivation of SBP coefficients. To provide justification for its effectiveness, simulation results are presented and compared with other benchmark precoding schemes.