Artificial Noise Aided Directional Modulation: A Transmitter Architecture Perspective

Directional modulation (DM), as a multi-antenna-based physical layer security (PLS) technology toward millimeter wave (mmWave) scenarios, is capable of guaranteeing effective and secure transmissions by simultaneously performing beamforming and artificial noise (AN). Over the past decade, the DM architecture has changed from a fully analog (FA) one to a fully digital (FD) one. Additionally, the hybrid architecture, which effectively balances both performance and hardware considerations, has unfortunately received limited attention regarding the integration of DM within its framework. In this paper, we intend to give a detailed study of DM technology under the still imperfect FA and more effective hybrid architectures. Specifically, for a single-user (SU) multiple-input single-output (MISO) FA-DM systems, considering the constant modulus constraint (CMC) of phase shifters (PSs) as well as the power allocation issue, we propose a low-complexity DM precoder with closed-form solutions by maximizing the entropy of precoder (EoP). On the other hand, optimization problems are cast and discussed under different RF chain number configurations for SU multiple-input multiple-output (MIMO) hybrid-DM systems. Specifically, as the RF chain count decreases from twice the data stream number to match it, we all use optimization to achieve the corresponding optimal or near-optimal solutions. Finally, simulation results verify our aforementioned analysis and demonstrate the superiority of the proposed FA and hybrid-DM algorithms in terms of secrecy rate and bit error rate (BER) compared to the counterparts when PSs with low resolution are used.