全球导航卫星系统由于信号微弱和结构公开等特点而易受干扰。其中，欺骗干扰可诱使接收机输出受攻击方控制的位置和时间信息，危害较大。尽管已有很多反欺骗技术的研究，但是大都处于欺骗信号检测、识别和抑制层面，只能让具备反欺骗能力的接收机免受欺骗。为了从源头清除欺骗威胁，还需对欺骗源进行定位。但是现有定位技术具有灵活性差、隐蔽性差、抗干扰能力弱、无法应对定向欺骗等不足，难以适应复杂电磁环境。为此，本文以复杂电磁环境为应用背景，针对军、民用户皆面临的转发式欺骗干扰，开展欺骗源定位技术研究。 首先，针对复杂电磁环境下的欺骗源定位问题，提出了一种基于单个接收机的空时双差(STDD)定位技术。区别于已有方法利用多个接收机，本文利用多颗卫星作为定位节点。由于卫星位置和时间同步信息可以从星历获取，通过提取转发式欺骗信号携带的欺骗源到卫星的距离信息可以对欺骗源进行定位。该距离信息含有信号达到时间、从欺骗源到接收机的传播时长和转发时延等未知参数。STDD定位技术通过空间、时间差分消除这些未知参数，使其不依赖接收机位置和时间同步信息。单个接收机仅接收欺骗信号即可定位欺骗源，适用于复杂电磁环境。基于该定位技术，根据欺骗源的不同运动状态分别实现了静止和运动欺骗源的定位。 另外，为了让空时双差定位技术能够在实际场景下落地，分别提出了基于接收机位移估计的欺骗检测技术、基于钟漂估计的欺骗识别技术和转发时延变化检测技术、基于欺骗源位移估计的欺骗源运动检测技术。欺骗检测技术能够检测到欺骗攻击是否发生。若发生欺骗，利用欺骗识别技术能够识别出哪些信号是欺骗信号，以确保定位所用的观测量均属于欺骗信号。转发时延变化检测技术能够解决复杂攻击模式下的欺骗源定位问题，以避免时延变化所引入的定位偏移。欺骗源运动检测技术能够解决不同场景下定位算法的选取问题，以避免静态定位算法用于动态场景导致的定位偏移和动态定位算法用于静态场景导致的性能损失。 最后，将上述技术融合得到完整的转发式欺骗源定位系统并开发了原理样机。利用样机分别对静止和运动欺骗源的定位性能进行了实验测试。实验结果表明，5秒观测数据时长下，4颗卫星时静态定位精度达到50至70米，8颗卫星时动态定位精度约100米。

The Global Navigation Satellite System is susceptible to interference including jamming and spoofing because of the weak power and open structure of the navigation signal. Among them, spoofing can induce a receiver to output wrong position and time information controlled by the spoofer, which is a huge threat. Although there have been many researches on anti-spoofing technologies, most of them are used for detection, identification and suppression of spoofing. As a result, only receivers equipped with these anti-spoofing technologies can be protected from spoofing. In order to remove the spoofing threat from the source, it is also necessary to locate the spoofer. However, the existing spoofer localization technologies are difficult to meet the requirements of complex electromagnetic environments, because they have poor flexibility, invisibility, and anti-interference ability, and cannot cope with the directional spoofing. As a result, this dissertation conducts the research on the localization of the replay-type spoofer named meaconer faced by both military and civilian users in complex electromagnetic environments. First of all, a space-time double-difference (STDD) method based on a single receiver is proposed in order to locate the spoofer in complex electromagnetic environments. The method utilizes multiple navigation satellites as positioning nodes to locate the spoofer, which is different from ideas of existing methods that uses multiple receivers. Since the satellite position and time synchronization information can be obtained from the satellite ephemeris, the proposed method locates the spoofer by extracting information about the distance between the spoofer and the satellite from the meaconing signal. The distance information contains unknown parameters such as the signal arrival time, the propagation time from the spoofer to the receiver, and the replay delay induced by the spoofer. The STDD method can eliminate these unknown parameters through space and time differences. Based on this, the method does not require the receiver position and time synchronization information. Besides, with this method a single receiver can locate the spoofer by only receiving spoofing signals, which is suitable for complex electromagnetic environments. On the basis of the STDD method, stationary and moving spoofers are located respectively according to their motion states. In addition, a spoofing detection method based on receiver displacement estimation, a spoofing identification method and a replay delay change detection method based on clock drift estimation, and a spoofer motion detection method based on spoofer displacement estimation are proposed respectively in order to implement the STDD method in actual spoofing scenarios. The spoofing detection method can detect whether spoofing is happening. If spoofing exists, the spoofing identification method can identify which signals belong to spoofing ones, which can ensure that the observations used for localization are all extracted from spoofing signals. The replay delay change detection method can cope with the localization of a spoofer that performs a complex spoofing attack, and avoid the positioning bias introduced by the delay change. The spoofer motion detection method is used to select the proper localization method for the stationary or moving spoofer. It can avoid the positioning bias caused by the static localization method, which is wrongly used to locate the moving spoofer. And it can also avoid the performance loss caused by the dynamic localization method, which is wrongly used to locate the stationary spoofer. Finally, the above methods are merged together in order to obtain a complete spoofer localization system and develop a principle prototype. Based on the prototype, the localization performance for the stationary and moving spoofers is evaluated through experiments, respectively. Experimental results show that the estimate accuracy for the stationary spoofer can achieve 50 to 70 meters with 4 satellites, and the estimate accuracy for the moving spoofer can achieve about 100 meters with 8 satellites when the observation duration is 5 seconds.