In this study, the authors study energy-efficiency (EE) optimisation of three-node multiple-input-multiple-output wiretap channels. In particular, they aim to maximise EE by optimising the transmit covariance under th...
详细信息
In this study, the authors study energy-efficiency (EE) optimisation of three-node multiple-input-multiple-output wiretap channels. In particular, they aim to maximise EE by optimising the transmit covariance under the constraint of transmit power consumption. Such optimisation problems are non-convex and difficult to solve. Here, they propose a novel optimisation algorithm, which is very different from previous approaches, with guaranteed convergence to the Karush-Kuhn-Tucker point. In particular, through a channel improvement argument and an intelligent bounding technique, they first obtain a min-max reformulation of their original problem. Then, based on the idea of alternative optimisation, they derive two sub-problems from it. To solve these two problems, they used the efficient barrier Newton and the gradient projection methods. The outcome from this last step is a semi-closed-form solution for the two sub-problems. The numerical results obtained from the proposed algorithm represent better performances than the existing methods.
暂无评论