Performance Analysis and Power Allocation for Downlink Cell-Free Massive MIMO System With Nonlinear Power Amplifier

The Cell-Free Massive Multiple-Input Multiple-Output (CF-mMIMO) system, incorporating nonlinear power amplifiers (PAs) at both access points (APs) and user equipment (UEs), is analyzed to ensure high and consistent data rates for sixth-generation (6G) networks while maintaining economic feasibility. Specifically, Minimum Mean Squared Error (MMSE) channel estimates and closed-form downlink achievable rates are derived under PA nonlinearity. We propose nonlinearity-aware power allocation algorithms to enhance performance. First, we present two baseline methods: Full Power and Channel Inversion. Subsequently, we develop two novel algorithms aimed at maximizing the sum rate (MSR) and the minimum rate (MMR). Using successive convex approximation (SCA) and single condensation method (SCM), we reformulate these optimization problems to convex subproblems. Geometric Programming (GP) is employed to address specific subproblems within them. Results show the proposed nonlinearity-aware power allocation algorithms outperform conventional linear PA algorithms, enabling CF-mMIMO to tolerate PA nonlinearity without performance loss. The MSR algorithm boosts sum throughput, minimum net throughput, and fairness, while the MMR algorithm enhances minimum rate, fairness, and average net throughput.