Exploiting Constructive Power Amplifier Nonlinearities through Symbol-Level Precoding

In this paper, we propose a symbol-level precoding (SLP) scheme in the presence of nonlinear power amplifiers (PAs) in multi-user multiple-input single-output (MU-MISO) downlink systems. Specifically, we address the distortion introduced by nonlinear PAs from the perspective of constructive interference (CI) by incorporating the memoryless polynomial model of nonlinear PAs into the SLP design, which can be exploited at symbol level. We firstly construct the optimization problem that aims to enhance the signal-to-interference-plus-noise ratio (SINR) without investing additional transmit power. Since the original optimization problem is non-convex, we introduce auxiliary variables to transform the optimization problem and adopt the alternating optimization framework for the new optimization problem. By employing several approximations, we transform the non-convex sub-optimization problems into their semidefinite programming (SDP) form, and obtain feasible solutions by the semidefinite relaxation (SDR) method. Numerical results verify that the proposed PA-aware SLP outperforms the state-of-the-art in terms of the error-rate performance in the presence of nonlinear PAs in the MU-MISO downlink.