摘要
研究了基于智能超表面的安全传输,现有研究主要针对窃听者CSI已知这一理想假设,而实际中窃听者CSI通常难以获取;因此,针对窃听者CSI完全未知条件下智能超表面的安全传输进行了研究;首先通过基站波束赋形向量和智能超表面相移矩阵的主被动波束赋形在满足合法用户通信质量的约束下最小化通信信号传输功率,然后将剩余功率用于发送人工噪声以干扰潜在的窃听者;提出了一种深度学习辅助的流形优化方法来解决这一功率分配问题,该方法将黎曼梯度下降模型与深度学习方法相结合,基于神经网络的自适应学习能力动态调控黎曼梯度下降的方向和步长;实验结果表明,与现有的优化算法相比,所提出的方法在达到几乎相同的安全速率的同时,计算复杂度降低至少一个数量级。
Research on the secure transmission based on reconfigurable intelligent surface(RIS)is conducted,where existing research mainly focuses on the ideal assumption of known eavesdropper's channel state information(CSI),while it is usually difficult for the eavesdropper's CSI to obtain in practice.Therefore,this paper studies a RIS-based secure transmission under completely unknown eavesdropper's CSI.Firstly,through the beamforming vectors in base stations and CIS phase shift matrices,the active-passive beamforming minimizes the transmission power of communication signals while meeting the constraints of legitimate users in communication quality,and then uses residual power to transmit artificial noise to jam potential eavesdroppers.A deep learning-based manifold optimization method is proposed to address the power allocation,which combines the Riemannian gradient descent model with deep learning method.The adaptive learning ability of neural networks can dynamically adjust the direction and step size of the Riemannian gradient descent.Experimental results show that compared to existing optimization algorithms,the proposed method reduces the computational complexity by at least one order of magnitude while achieving almost identical secrecy rate.
作者
苗睿锴
宋志群
李勇
李行健
刘丽哲
王斌
MIAO Ruikai;SONG Zhiqun;LI Yong;LI Xingjian;LIU Lizhe;WANG Bin(The 54th Research Institute,China Electronics Technology Group Corporation,National Key Laboratory of Advanced Communication Networks,Shijiazhuang 050081,China)
出处
《计算机测量与控制》
2025年第12期286-295,共10页
Computer Measurement & Control
基金
河北省自然科学基金(F2024523005)
河北省博士后科学基金(B2023005001)
先进通信网全国重点实验室基金(FFX24641X005)
通信抗干扰全国重点实验室基础科研创新基金(稳定支持)项目(IFN202404)。
