Saturation Power based Simple Energy Efficiency Maximization Schemes for MU-MISO Systems

TL;DR

This paper introduces a low-complexity method for maximizing energy efficiency in MU-MISO systems by leveraging saturation power, which simplifies the problem to spectral efficiency maximization below a certain power threshold.

Contribution

The paper proposes a novel approach to compute saturation power using channel statistics, enabling efficient EE maximization with reduced computational complexity.

Findings

Achieves near-optimal EE performance

Reduces computational complexity significantly

Provides insight into saturation power behavior

Abstract

In this paper, we investigate an energy efficiency (EE) maximization problem in multi-user multiple input single output downlink channels. The optimization problem in this system model is difficult to solve in general, since it is in non-convex fractional form. Hence, conventional algorithms have addressed the problem in an iterative manner for each channel realization, which leads to high computational complexity. To tackle this complexity issue, we propose a new simple method by utilizing the fact that the EE maximization is identical to the spectral efficiency (SE) maximization for the region of the power below a certain transmit power referred to as saturation power. In order to calculate the saturation power, we first introduce upper and lower bounds of the EE performance by adopting a maximal ratio transmission beamforming strategy. Then, we propose an efficient way to compute the saturation power for the EE maximization problem. Once we determine the saturation power corresponding to the maximum EE in advance, we can solve the EE maximization problem with SE maximization schemes with low complexity. The derived saturation power is parameterized by employing random matrix theory, which relies only on the second order channel statistics. Hence, this approach requires much lower computational complexity compared to a conventional scheme which exploits instantaneous channel state information, and provides insight on the saturation power. Numerical results validate that the proposed algorithm achieves near optimal EE performance with significantly reduced complexity.