信道不确定下的窃听者鲁棒性优化算法研究

摘要

This paper considers a wireless communication system where the source node sends messages with the help of relay nodes to the destination node in the presence of an eavesdropper. This paper assumes that the channel state information (CSI) of the eavesdropper is not perfect, but channel estimation error satisfies Gaussian distribution. With the available imperfect channel estimate, we propose the robust model for the signal-to-interference-plus-noise (SINR) of the eavesdropper. Firstly, we transform the problem of the worst case SINR to a norm bounded error model which can be relaxed into semi-definite programming problem (SDP), and we can obtain the rank-one solution of SDP prob-lem. Secondly, for the communication quality protection of our system, we get an optimization model which minimizes the total relay power consumption subject to the SINR of the intended destination above a certain threshold and the worst case SINR of the eavesdropper below another threshold, and we can relax the problem into a SDP in the same way and find the rank-one solution of SDP problem. Finally, numerical results are provided to prove the effectiveness of the proposed algorithm.%本文研究在存在一个窃听者情况下信源在中继辅助下向目的节点传输信号的单向通信系统。本文假定窃听者信道状态信息部分已知，信道估计误差服从高斯分布。在此情况下，本文提出了针对窃听者信干噪比（SINR）的鲁棒性设计算法，首先将窃听者最糟糕 SINR 问题转换成范数有界模型，并进一步松弛成带两个约束的半定规划问题（SDP），求得其秩一解。然后针对通信系统的通信质量保护，本文建立了在目的节点 SINR 大于某一阈值和窃听者最糟糕 SINR 小于某一阈值约束下，中继消耗功率最小为目标的优化模型，同样将问题转化成 SDP，求得原问题的最优解。最后，仿真结果证明了该算法的有效性。