Secrecy Analysis of Cell-free Massive MIMO-enabled UAV Communications

This paper studies physical layer security for cell-free (CF) massive multiple-input-multiple-output (mMIMO)-enabled unmanned aerial vehicles (UAV) communication under threats from multiple UAV eavesdroppers (UAV-Eves). We consider spatially correlated Rician fading channels and two eavesdropping modes: non-colluding and colluding UAV-Eves. To counter multi-antenna UAV-Eves with successive interference cancellation capabilities, we propose an artificial noise (AN)-aided secure transmission scheme. Incorporating active pilot attacks and user-centric (UC) clustering, we derive a closed-form secrecy rate lower bound under conjugate beamforming. Numerical results show that UC clustering significantly improves security compared to CF architecture, whereas pilot attacks degrade secrecy performance. Moreover, colluding UAV-Eves exhibit substantially higher security threats than non-colluding ones. The proposed AN-aided scheme ensures robust secrecy enhancement, particularly against advanced UAV-Eves.