K. Zarrinnegar, S. Tohidi, M. R. Mosavi, A. Sadr, D. M. de Andrés,
Volume 19, Issue 1 (3-2023)
Abstract
The Global Positioning System (GPS) is vulnerable to various deliberate and unintentional interferences. Therefore, identifying and coping with various interferences in this system is essential. This paper analyzes a method of reducing the dimensions of Cross Ambiguity Function (CAF) images in improving the identification of spoofing interference at the GPS using Multi-Layer Perceptron Neural Network (MLP NN) and Convolutional Neural Network (CNN). Using the proposed method reduces data complexity, which can reduce the number of learning data requirements. The simulation results indicate that, by applying the proposed image processing algorithm for different dimensions of CAF images, the CNN performs better than MLP NN in terms of training accuracy; the MLP NN is superior to CNN in terms of convergence speed of training. In addition, the results demonstrate that the operation of the proposed method is appropriate in the case of small-delay spoofed signals. Therefore, for the intervals above 0.25 code chip, the proposed method detects spoofing attacks with a correct detection probability close to one.
M. J. Jahantab, S. Tohidi, Mohammad Reza Mosavi, Diego Martín de Andrés,
Volume 22, Issue 0 (3-2026)
Abstract
Global Positioning System (GPS) spoofing poses serious threats to navigation systems, as it transmits false GPS signals that cause receivers to compute incorrect positions. To address this issue, our research in this study focused on leveraging the Cross-Ambiguity Function (CAF) along with advanced machine learning techniques to effectively detect spoofing attacks. A further challenge in using CAF for spoofing detection is its high dimensionality, which demands powerful hardware and considerably slows down the detection process. Detecting spoofing signals with delays of less than 0.5 chips relative to the authentic signal is particularly difficult. To overcome this, the SVD_Var dimensionality reduction algorithm, which leverages the variance of CAF data through Singular Value Decomposition (SVD), is proposed to enhance both speed and detection performance. The reduced-dimensionality data are subsequently used to train a basic Multi-Layer Perceptron (MLP) neural network and the k-Nearest Neighbors (kNN) algorithm. The effectiveness of the proposed method is validated using the widely recognized Texas Spoofing Test Battery (TEXBAT) dataset. Results indicate that the method achieves an average detection rate exceeding 80% across various TEXBAT scenarios, demonstrating enhanced sensitivity and robustness in spoofing detection compared to both traditional and state-of-the-art approaches. Also, this approach accomplishes a dimensionality reduction ranging from 99.69% to 99.99% in terms of the number of pixels which significantly accelerates the processing speed.
