Impact of Imperfect CSI on Projection-Based Secure MIMO Communications
This paper investigates physical-layer security in multiple-input multiple-output (MIMO) systems under imperfect channel state information (CSI). A unified framework is proposed, encompassing both cell-free and single-cell massive MIMO architectures, where CSI errors are modeled as Gaussian-distributed perturbations. The study focuses on projection-based secure precoding schemes applied to minimum mean square error (MMSE) and zero-forcing (ZF) precoders. It is shown that the resulting secure precoder can be decomposed into two components: one orthogonal to the eavesdropper’s channel subspace and another representing colored noise that characterizes the effect of CSI imperfections and constrains secrecy. Theoretical analysis demonstrates that the ZF-based projected-channel precoder preserves the ZF property under small CSI errors, thus exhibiting strong robustness. Simulation results in realistic multipath environments confirm the analysis, showing significant secrecy rate gains under CSI uncertainty (up to $\approx 15$ bps/Hz in high channel correlation configuration for the ZF-based projected channel scheme at SNR of 25 dB), over state-of-the-art approaches such as SVD-based and artificial-noise injection schemes.