Optimal semiadaptive transmission with artificial-noise-aided beamforming in MISO wiretap channels

This paper studies the design of artificial-noise-aided (AN-aided) secure transmission in slow-fading multiple-input-single-output (MISO) channels. We consider the scenario where a multiantenna transmitter sends messages to a single-antenna legitimate receiver in the presence of multiple noncolluding singleantenna eavesdroppers. It is assumed that the transmitter possesses instantaneous channel state information (CSI) of the legitimate channel but only the statistics of the eavesdropper's channels. Instead of adopting the well-known adaptive and nonadaptive transmission schemes, we consider the newly proposed semiadaptive scheme, in which the communication rate is fixed, and the secrecy rate is adaptively adjusted according to the legitimate channel's CSI. Compared with the adaptive scheme, a major advantage of fixing the communication rate is that the codebook can be fixed, and thus, the encoder and decoder designs are both simplified. We consider a generalized AN-aided beamforming structure, which allows arbitrary beamforming direction. Specifically, we focus on the problem of maximizing the secrecy throughput on the premise of reliable and secure transmission. By solving this problem, the optimal beamforming direction, powerallocation parameters, and secrecy rate are derived in closed form, and the optimal communication rate is obtained numerically. Finally, simulation results are presented to show the performance of the semiadaptive scheme.