Robust energy-efficient power loading for MIMO system under imperfect CSI

In this paper, we will investigate the energy efficient power loading problem in multiple-input multiple-output singular value decomposition (MIMO-SVD) architecture. Most existing power loading schemes are developed on the assumption that a scheduler possesses perfect channel stateinformation (CSI). However, in practice, perfect CSI is not always available. In this paper, we propose an energy-efficient power loading scheme for MIMO system under imperfect CSI, taking into account the effects of channel estimation error (CEE). Specifically, we propose two algorithms to maximise the system energy efficiency. One is a global method which can achieve the global optimal solution with high computational complexity; the other is a suboptimal method which can obtain the approximate optimal solution with lower computational complexity. Simulation results validate the effectiveness and robustness of our proposed power loading scheme.