Artificial Noise Aided Energy Efficiency Optimization in MIMOME System With SWIPT

This letter studies energy efficient optimization problem for a general multiple-input multiple-output multiple-eavesdropper system, in which the legitimate receiver exploits a power splitting (PS) scheme for decoding information and harvesting energy simultaneously. In order to cripple the eavesdroppers' reception and provide energy for the legitimate receiver to harvest, the transmitter also emits artificial noise (AN) in the transmission. Our interest in this letter lies in the joint design of the transmit covariance matrices and PS ratio, such that the secrecy energy efficiency (SEE) is maximized. The SEE maximization problem is a nonconvex fractional problem. To handle it, we resort to the fractional programming to transform its objective function into a more tractable subtractive form. Furthermore, we show that the resultant problem is primal decomposable by recasting it as an equivalent form amenable to alternating optimization. Numerical results indicate that incorporating AN and PS is beneficial to the SEE enhancement.