LAUSR

In this paper, we investigate the worst-case robust secure precoding design for simultaneous wireless information and power transfer in multiple-input-multiple-output wiretap system, which consists of one transmitter, one information receiver, and one energy receiver. By treating the ER as potential eavesdropper, we aim to maximize the worst-case secrecy rate by designing the transmit precoding matrix. This is a typical non-convex optimal design while the channel state information uncertainties make this problem harder to handle, thus we propose a new method to obtain the solution. Specifically, instead of approximating the logarithmic determinant as a trace, we propose to linearize the two log-det terms. After linearization, epigraph reformulation is used to deal with the bounded channel uncertainties. Then, an alternating optimization method is utilized to solve the reformulated problem. After obtaining the precoding matrix, we derive the worst-case secrecy rate by solving another optimal problem with respect to the channel uncertainty. Numerical results validate the performance of our proposed design.