基于半定规划方法的多个窃听用户认知网络物理层安全优化设计

Improved Transmit Design for Physical Layer Security in Cognitive Radio Networks with Multiple Eavesdropper Base on Semi-definite Programming

针对具有多个多天线窃听者的认知无线电网络(CRN),为了使系统保密速率达到最大,该文把对主用户的干扰设计成为一个约束条件,通过对次用户发送端传输协方差矩阵的优化设计来提高物理层安全性能。在已知信道状态信息(CSI)时利用矩阵性质和Charnes-Cooper变换,将该非凸函数转化为一个半定规划(SDP),从而得到次用户发送端的优化方案。仿真结果表明,相对于现有二次优化传输策略,该方案能够使系统的保密速率更大,并在复杂度方面具有优势。

In the Cognitive Radio Network（CRN） with multiple multi-antenna eavesdroppers, to make the system security rate maximum, secure communication over the physical layer that is subjected to the interference power constraints at the Primary Users（PU） is provided by designing the transmit covariance optimization of Secondary User Transmitter（SU-Tx）.When the Channel State Information（CSI） is known, the properties of the matricesand Charnes-Cooper transformation are used, the non-convex function is converted to a Semi-Definite Programming（SDP） to get the optimization scheme of SU-Tx. Simulation results show that compared with the existing sub-optimal transmission designs, the proposed method improves the secrecy rate and has more advantages on the complexity.