Distributed Precoding for Mitigating Hardware Impairments in Cell-Free Massive MIMO with Sequential Fronthaul

This paper addresses the challenge of mitigating hardware impairments (HWIs) in cell-free massive MIMO (CF-mMIMO) networks while maintaining scalability. Our proposed solution incorporates a hardware-aware sequential precoding consisting in the use of a serial fronthaul topology to exchange channel state information (CSI) and estimated hardware distortions between access-points (APs) to perform an over-the-air cancellation of the HWI. Each AP computes its local spatial multiplexing weights and then aims at transmitting the inverse of the HWI delivered by the preceding AP of the chain using the shared information. Furthermore, we adopt the Partial Zero-Forcing (PZF) precoding strategy, which leverages differences in user channel conditions to classify users into a strong and a weak group based on large-scale fading coefficients. Strong users are served with Zero-Forcing precoding to minimize interference, while weak users benefit from Maximum Ratio Transmission (MRT). The HWI compensation is applied only on strong users, ensuring efficient resource allocation and reduced complexity while offering better performance than the use of Full-Pilot Zero forcing (FZF). Simulation results demonstrate that our PZF-based sequential hardware-aware precoder significantly improves spectral efficiency even under high HWI conditions, making it a practical and robust solution for next-generation CF -mMIMO networks.