GPS Spoofing Detection Based on Decision Fusion with a K-out-of-N Rule

In order to obtain higher detection probability of the GPS spoofing, a general identification scheme with decision fusion is proposed. Firstly, the singular values of the wavelet transformation coefficients of both spoofing and genuine signal are computed and formed as the feature vectors. Then, the feature vectors are input into three classifiers, which are the support vector machines (SVM), the probabilistic neural networks (PNN) and the decision tree (DT), respectively, for GPS spoofing identification. Finally, the results of the three classifiers are fused with a K-out-of-N decision rule, and the final classification result is obtained. Simulation results exhibit the effectiveness of the proposed scheme, whose detection probability has increased by 3.75%, 5.06% and 12.36% than that of the SVM, the PNN and the DT on average, respectively. Moreover, the false alarm probability of the proposed scheme is lower than that of the three classifiers. In addition, the area under curve (AUC) is given to verify the effectiveness and feasibility of the proposed method.