Energy Efficient Artificial Noise-Aided Precoding Designs for Secured Visible Light Communication Systems

Abstract

Physical layer security (PLS) has recently gained a lot of attention in the research and development of visible light communication (VLC). In this article, we study the designs of PLS in VLC systems in the presence of multiple unauthorized users (i.e. eavesdroppers) using artificial noise (AN)-aided precoding. The design objective focuses on minimizing the total transmit power subject to specific constraints on the signal-to-interference-plus-noise ratios (SINRs) of the legitimate and unauthorized users. In particular, two design approaches are investigated considering the availability of unauthorized users' channel state information (CSI) at the transmitter. In the case of unknown CSI, the AN is constructed to lie on the null-space of the legitimate user's channel. The design problem is convex, thus, can be effectively solved. When the CSI is available, the design additionally imposes constraints on the maximum allowable unauthorized users' SINRs. The design problem, in this case, is, nevertheless, non-convex. Therefore, instead of finding the optimal solution, we examine two different sub-optimal yet low-complexity approaches to solve the problem, namely: Concave-Convex Procedure (CCP) and Semidefinite Relaxation (SDR). Additionally, robust designs that take into account channel uncertainty are also investigated. Extensive numerical results are shown to demonstrate the feasibility and performance of each design with practical parameters.