Radio Frequency Interference (RFI) detection and characterization play a critical role in ensuring the security of all wireless communication networks. Advances in Machine Learning (ML) have led to the deployment of many robust techniques dealing with various types of RFI. To sidestep an unavoidable complicated feature extraction step in ML, we propose an efficient Deep Learning (DL)-based methodology using transfer learning to determine both the type of received signals and their modulation type. To this end, the scalogram of the received signals is used as the input of the pretrained convolutional neural networks (CNN), followed by a fully-connected classifier. This study considers a digital video stream as the signal of interest (SoI), transmitted in a real-time satellite-to-ground communication using DVB-S2 standards. To create the RFI dataset, the SoI is combined with three well-known jammers namely, continuous-wave interference (CWI), multi- continuous-wave interference (MCWI), and chirp interference (CI). This study investigated four well-known pretrained CNN architectures, namely, AlexNet, VGG-16, GoogleNet, and ResNet-18, for the feature extraction to recognize the visual RFI patterns directly from pixel images with minimal preprocessing. Moreover, the robustness of the proposed classifiers is evaluated by the data generated at different signal to noise ratios (SNR).
Although the Global Navigation Satellite System (GNSS) technology provides an excellent benefit in different critical areas such as civilian, aviation, military, and commercial applications, it is highly vulnerable to various signal disruptions causing significant positioning errors. One of the major threats to a GNSS receiver is the intentional interference known as jamming. A Jammer significantly disrupts the normal functioning of a GNSS receiver, at the acquisition, tracking, and positioning stages. The foremost important step to combat against jamming of GNSS signals is the early detection and characterization of the interfering signals to guarantee the Quality of Service (QoS). This paper presents a robust Deep-Learning (DL) based technique using transfer learning to characterize the type of disruption in GNSS signal based on time-frequency analysis. To this end, a pretrained Convolutional Neural Network (CNN) is used to extract the informative features from the scalogram of the received signals. Further, a fully connected layer followed by a Soft-Max activation function is deployed to classify the signals. In this work, the Signal of Interest (SoI) is a synthetic GPS signal generated by a GNSS simulator. In our experiment, the GPS signal is combined with different kinds of jamming, spoofing, and multipath signals. Moreover, the proposed classification approach can recognize not only the various kinds of jammers such as ones producing Continuous Wave Interference (CWI), Multi-CWI (MCWI), Chirp Interference (CI), and Pulse interference (PI) but also the inclusion of Additive White Gaussian Noise (AWGN). Besides that, the effect of five pre-trained CNNs, namely, AlexNet, GoogleNet, ResNet-18, VGG-16, and MobileNet-V2, is evaluated on classification accuracy. The GNSS signal and its seven disruptive variants are recorded at three different power levels such as low, medium, and high. The medium power level signal is used for training and the testing has been carried out for unseen data set of low, high, and mixed power level. From the simulation results, it has been observed that MobileNet-V2 has performed better than other techniques with an accuracy of 99.8%. Finally, the trained MobileNet-V2 is used to predict the unseen data type generated at different Jamming to signal Ratios (JSRs).
Radio Frequency Interference (RFI) detection and characterization play a critical role to in ensuring the security of all wireless communication networks. Advances in Machine Learning (ML) have led to the deployment of many robust techniques dealing with various types of RFI. To sidestep an unavoidable complicated feature extraction step in ML, this paper proposes an efficient end-to-end method using the latest advances in deep learning to extract the appropriate features of the RFI signal. Moreover, this study utilizes the benefits of transfer learning to determine both the type of received RFI signals and their modulation types. To this end, the scalogram of the received signals is used as the input of the pre-trained convolutional neural networks (CNN), followed by a fully-connected classifier. This study considers a digital video stream as the signal of interest (SoI), transmitted in a real-time satellite-to-ground communication using DVB-S2 standards. To create the RFI dataset, the SoI is combined with three well-known jammers namely, continuous-wave interference (CWI), multi- continuous-wave interference (MCWI), and chirp interference (CI). This study investigated four well-known pre-trained CNN architectures, namely, AlexNet, VGG-16, GoogleNet, and ResNet-18, for the feature extraction to recognize the visual RFI patterns directly from pixel images with minimal preprocessing. Moreover, the robustness of the proposed classifiers is evaluated by the data generated at different signal to noise ratios (SNR).
No abstract
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.