محمدرسول طاهري

Abstract Increasing reliance on Global Positioning System (GPS) technology in various sectors has exposed vulnerabilities to malicious attacks, particularly GPS jamming and spoofing. This thesis deals with the analysis of threat detection and mitigation strategies to increase the resistance of GPS receivers against jamming attacks. This thesis investigates and develops a method based on deep learning to detect and classify jamming in GPS signals. Jamming, as one of the important challenges in GPS systems, can lead to disruption in the correct functioning of positioning and navigation. The main goal of this research is to design and implement a model that can distinguish normal GPS signals from signals affected by jamming and, in the next step, classify different types of jamming. To achieve this goal, convolutional neural networks (CNN) and transfer learning techniques have been used. The input data to the model was extracted in the form of signal spectrograms and the model was trained using these features. The performance of the model was investigated in different conditions with different jamming to signal ratios (JSR) in order to eva‎luate its accuracy in identifying and classifying types of jamming at different power levels. The results of this research show that the developed model has been able to distinguish jammed signals from normal signals with high accuracy, even in the conditions where the jamming power is low compared to the signal. However, the accuracy of the model in classifying different types of jamming has decreased with the decrease of JSR. This indicates the need to improve the model and use more advanced methods to increase classification accuracy in low noise conditions. Overall, this research is an effective step towards applying deep learning techniques to deal with GPS jamming threats, and its results can bring significant improvements in future positioning and navigation systems. Key words: GPS jamming, deep learning, convolutional neural network (CNN), transfer learning, signal spectrogram, jamming detection, jamming classification, jamming to signal ratio (JSR), positioning systems.

تشخيص جمينگ , طبقه بندي جمينگ , يادگيري ماشين , يادگيري عميق

Jamming detection , Classification of jamming , ِDeep learning , machine learning

MohammadRasool Taheri

Mahdi NasiriSarvi