中文摘要：

如何抗欺骗干扰是卫星导航系统安全通信的保障。针对欺骗干扰问题，提出了基于自适应增量流形学习的GNSS欺骗干扰识别方法。流形学习作为非线性特征提取方法，能够有效提取欺骗干扰信号与真实信号的内在特征，但面对持续的数据流时并不能实时进行特征提取。因此，提出一种自适应选择重构邻域的增量流形学习算法，能够实现在线提取特征，并解决邻域构建问题，最后利用支持向量机（SVM）实现干扰识别。仿真实验分析不同情况下的识别性能，表明该算法能够提高欺骗干扰识别效率，并能取得很好的识别效果。

英文摘要：

How to deal with anti-spoofing is the guarantee for security communication of the GNSS （global satellite navigation system）. Aiming at GNSS spoofing identification, a new method based on manifold learning is proposed. Manifold learning, as a nonlinear method for feature extraction, can effectively extract the intrinsic characteristics of the spoofing signal and the real signal. Additonally, in order to apply manifold learning in reahime, an incremental manifold learning algorithm of self-adaptive selective reconstruction neighborhood is proposed. Based on resolution of neighborhood construction, the incremental problem of manifold learning is solved, and the learning efficiency is improved. Finally, SVM（support vector machine） is used to realize the interference identification. Simulations indicate that the proposed approach could improve the efficiency of GNSS spoofing identification and acquire better recognition performance.