Satellite navigation is widely used in daily life, and spoofing poses a serious security threat to satellite navigation systems. Most of the current spoofing detection algorithms are designed based on the receiver-solving process, which cannot be applied to factory receivers. In the case of single station spoofing, according to the spoofing signal, the successfully spoofed receivers output the same location information. Due to the different receiver antenna positions, the clock offsets of each receiver are different and positively correlated with the receiver antenna distance. Based on the above principles, this paper proposes a multi-receiver spoofing detection algorithm. Firstly, the spoofing attack is preliminarily judged by the baseline length between antennas. Then the receiver position vector matrix is constructed through multiple receiver position information, the position vector matrix is decomposed by Singular Value Decomposition (SVD) method, and the redundant information is detected by Principal Component Analysis (PCA) method ，and determine whether the receiver is spoofed by a spoof detector. The experimental results show that the position vector matrix is linearly correlated when there is a spoofing attack, and the eigenvalue energy is concentrated in the first receiver position vector; when there is no spoofing signal, the position vector matrix is linearly independent, and the eigenvalues are uniformly distributed. This method can simply and effectively detect whether there is a spoofing attack on the receiver.