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

Physical layer security (PLS) has gained a lot of attention in the research and development of visible light communication (VLC) over the past few years. In this paper, we consider improving PLS in VLC systems with the presence of multiple eavesdroppers by means of artificial noise (AN)-aided precoding. Our focus is on minimizing the average total transmit power of the information-bearing signal and the AN subject to specific constraints on the signal-to-interference-plusnoise ratios (SINRs) of the legitimate user and eavesdroppers. In particular, the design aims at guaranteeing a minimum threshold on the legitimate user’s SINR while limiting the eavesdroppers’ SINRs to a predefined maximum allowable value. The optimal design problem is shown to be non-convex, thus it is difficult to optimally solve in reasonable time. Instead of finding the optimal solution, we investigate two different sub-optimal yet low-complexity approaches to solve the problem, namely: Successive Convex Approximation (SCA) and Semidefinite Relaxation (SDR). Numerical results are presented to illustrate the efficiency of the two techniques.