JPC-GCE Algorithm for Energy-Efficient Wideband mmWave Systems

In this paper, a joint precoding and combining (JPC) algorithm based on generalized cross-entropy (GCE) optimization is proposed for energy-efficiency (EE) wideband millimeter wave (mmWave) systems, named as JPC-GCE algorithm. Firstly, an analog precoder and combiner (PCr) joint optimization problem is converted to a high-quality PCr sample selection problem based on probability distribution. Then, the quantized phase probability distribution parameters are updated by the GCE and Lagrange multiplier methods. Via this, the feasible set of optimal analog PCr vectors can be obtained. Especially, to improve EE, a low-resolution phase shifter (PS) is used in the analog PCr design, which reduces power consumption by PS quantization. Moreover, to eliminate the impact of interference on spectral efficiency (SE) performance, the digital precoder can be simply computed by the block diagonal approach, after selecting the most effective analog PCr for each subcarrier. Simulation results show that the JPC-GCE algorithm not only has high EE by selecting low-resolution PS while achieving better SE.