Beamforming for MIMO Gaussian wiretap channels with imperfect channel state information

In this paper, we propose a new beamforming scheme for multi-input multi-output (MIMO) Gaussian wiretap channels where the channel state information (CSI) from the eavesdropper is imperfectly known to the transmitter. A stochastic model is constructed to characterize the imperfect CSI of the eavesdropper, in which a factor 0 ≤ τ ≤ 1 is introduced to describe the degree of the available eavesdropper's channel knowledge at the transmitter. When τ varies from 0 to 1, the eavesdropper's channel knowledge available at the transmitter ranges from statistically known to perfectly known. We design the proposed beamforming scheme by maximizing a lower bound on the achievable secrecy rate. We first demonstrate that our scheme achieves higher secrecy rate than the existing eigenvalue decomposition-based beamforming scheme which is optimal for τ = 0. We then demonstrate that the proposed scheme achieves higher secrecy rate than the existing generalized eigenvalue decomposition-based beamforming scheme which is optimal for τ = 1. Furthermore, we derive tight approximations for the proposed beamforming scheme in the high signal-to-noise ratio (SNR) regime and the low SNR regime. The accuracy of these approximations is validated via numerical results. Finally, we demonstrate that our proposed scheme achieves almost the same secrecy performance as the optimal beamforming solution that is obtained through numerical search.