Hybrid digital to analog (D-A) precoding is widely used in millimeter wave systems to reduce the power consumption and implementation complexity incurred by the number of radio frequency (RF) chains that consume a lot of the transmitted power in this system. In this paper, an optimal number of RF chains is proposed to achieve the desired energy efficiency (EE). Here, the optimization problem is formulated in terms of fractional programming maximization, resulting in a method with a twofold novelty: First, the optimal number of RF chains is determined by the proposed bisection algorithm, which results in an optimized number of data streams. Second, the optimal analog precoders/combiners are designed by eigenvalue decomposition and a power iteration algorithm, followed by the digital precoders/combiners which are designed based on the singular value decomposition of the proposed effective uplink and downlink channel gains. Furthermore, the proposed D-A systems are designed carefully to attain a lower complexity than the existing D-A algorithms while achieving reasonable performance. Finally, the impact of utilizing a different number of quantized bits of resolution on the EE is investigated. Simulation results show that the proposed algorithms outperform existing algorithms in terms of EE, spectral efficiency, and computational complexity.
|Number of pages
|IEEE Transactions on Green Communications and Networking
|Early online date
|7 Feb 2020
|Published - 1 Jun 2020