DETECTION OF VLF TRANSMITTERS USING THE ViTRANSFORMER DEEP LEARNING ALGORITHM

АУТОР(И) / AUTHOR(S): Danilo Lazović , Olivera Pronić-Rančić , Aleksandra Nina , Jovan Bajčetić 

Download Full Pdf   

DOI:  10.46793/Geoscira25.073L

САЖЕТАК / ABSTRACT:

Radio signals of very low frequency (VLF) are used for monitoring the lower ionosphere. Their analysis provides information about ionospheric perturbations in the areas between the transmitter and the receiver of the observed signal, caused by phenomena in space and within Earth’s layers. The research presented in this paper demonstrates the capabilities of the ViTransformer neural network for detecting VLF transmitters using images. The images used in this study represent a two-dimensional visualization of the signal emitted by a VLF transmitter and received by a corresponding receiver. For training and testing, the data recorded at the Institute of Physics in Belgrade were used. The database employed contains recorded signals emitted from three different VLF transmitters located in different parts of the world (Germany, the United States, and Australia). The research results show that using the ViTransformer neural network makes it possible to perform classification with high accuracy, reaching nearly 95%.

КЉУЧНЕ РЕЧИ / KEYWORDS:

ViTransformer, detection, VLF transmitter

ПРОЈЕКАТ / ACKNOWLEDGEMENT:

The authors acknowledge funding provided by the Institute of Physics Belgrade through the grants by the Ministry of Science, Technological Development and Innovation of the Republic of Serbia.

ЛИТЕРАТУРА / REFERENCES:

• Silber, I., and Price, C. (2017). On the Use of VLF Narrowband Measurements to Study the Lower Ionosphere and the Mesosphere–Lower Thermosphere. Surv. Geophys., 38, 407.
• Seybold, J. S. (2005). Introduction to RF Propagation. John Wiley and Sons, New Jersey, USA.
• Lai, Y. (2019). A Comparison of Traditional Machine Learning and Deep Learning in Image. J. Phys.: Conf. Ser., 1314, 012148
• Vaswani, A., Shazeer, N., Parmar, N. et al. (2017). Attention Is All You Need. 31st Conference on Neural Information Processing Systems, Long Beach, CA, USA.
• Vakanski, A. (2023). Lecture 20 – Transformer Networks. https://fall-2023-python-programming-for-data-science.readthedocs.io/en/latest/Lectures/Theme_3-Model_Engineering/Lecture_20-Transformer_Networks/Lecture_20-Transformer_Networks.html#20.9-Vision-Transformers.
• Boudjada, M. Y., Biagi, P. F., Eichelberger, H. U. et al. (2024). Analysis of Pre-Seismic Ionospheric Disturbances Prior to 2020 Croatian Earthquakes. Remote Sens., 16, 529.
• Nina, A. (2024). Analysis of VLF Signal Noise Changes in the Time Domain and Excitations/Attenuations of Short-Period Waves in the Frequency Domain as Potential Earthquake Precursors. Remote Sens., 16, 397.
• Nina, A., Pulinets, S., Biagi, P. F. et al. (2020). Variation in natural short-period Ionospheric noise, and acoustic and gravity waves revealed by the amplitude analysis of a VLF radio signal on the occasion of the Kraljevo earthquake (Mw = 5.4). Sci. Total Environ., 710, 136406.
• Nina, A., Radovanović, M., Milovanović, B. et al. (2017). Low ionospheric reactions on tropical depressions prior hurricanes. Adv. Space Res. 60, 1866.
• Price, C., Asfur, M., Yair, Y. et al. (2009). Maximum hurricane intensity pre-ceded by increase in lightning frequency. Nat. Geosci., 2, 329.