Secrecy Performance Analysis for An-Aided Linear ZFBF in MU-MIMO Systems with Limited Feedback

Although there have been extensive works on artificial-noise-aided (AN-aided) secure transmission schemes for multi-antenna systems, there is still lack of study on the AN-aided scheme employing the widely used linear zero-forcing beamforming (ZFBF) for systems with multiple distributed users. Particularly, there is no analytical secrecy performance for general system settings with neither perfect nor imperfect channel state information of the legitimate users at the transmitter (CSIT). This paper considers the AN-aided ZFBF based on the quantized CSIT for secure communication in the downlink multiuser multi-antenna systems with an external multi-antenna eavesdropper. We develop an approximated closed-form lower bound on the ergodic rate of each legitimate user (LU), and also a closed-form upper bound on the maximum achievable ergodic rate for each LU's messages over the eavesdropper's channel without assuming any asymptotes for system parameters. Then, an approximated closed-form lower bound on the ergodic secrecy rate of each LU follows. Simulation results validate our analytical secrecy performance results and also the effectiveness of AN in enhancing the secrecy performance of linear ZFBF.