Robust artificial noise aided transmit designs for physical layer security

For the cases of imperfect channel state information (CSI) on the main channel in the physical-layer-security communications, we study the artificial-noise-aided robust transmit designs. We consider two optimization problems for the multi-input single-output single-antenna eavesdropper (MISOSE) and multi-input multi-output multi-antenna eavesdropper (MIMOME) systems. Under the deterministic uncertainty model (DM), the optimization problem aims to minimize the total transmit power while guaranteeing the probabilistic constraint for the signal to interference and noise ratios (SINR) at the eavesdropper, and the SINR constraint at the legitimate receiver. For the stochastic uncertainty model (SM), the optimization problem is also to minimize the total transmit power under the probabilistic constraints for the SINRs at the legitimate receiver and the eavesdropper. These problems are hard to be solved directly. We transform the optimization problems into a single-variable optimization problem with solving a sequence of semidefinite programs. By comparison, we propose another low-complexity method based on the expectation value. Finally, we verify the performance for the proposed robust transmit designs by the simulation results.