Novel modulation recognizer for frequency-hopping signals based on persistence diagram
Document Type
Conference Proceeding
Publication Date
10-27-2020
Abstract
Modulation recognition or classification has drawn a lot of research interest for military and civilian communication applications. However, modulation recognition (classification) for frequency-hopping (FH) signals remains very difficult due to its unique time-varying spectral characteristics. In this work, a novel modulation-recognition scheme using the topological data analysis is proposed as the first attempt to solve this difficult problem. In our proposed scheme, the received frequency-hopping signal is sampled and these samples are mapped into the embedded phase space. Then the expected persistence diagrams (EPDs) are established to construct the features. We treat such resulted EPDs as probability distribution functions. Consequently, the discrepancy between any two EPDs can be measured by Kullback-Leibler divergence (KLD). During the recognition stage, a test signal will thus be converted to an EPD and the corresponding KLDs to all training features (EPDs) can be calculated for modulation recognition decision. Preliminary results from Monte Carlo simulations have demonstrated that the recognition accuracy of our proposed new approach is above 95% when the signal-to-noise ratio is greater than 30 dB.
Publication Source (Journal or Book title)
IEEE International Symposium on Broadband Multimedia Systems and Broadcasting, BMSB
Recommended Citation
Huang, X., Zhang, L., Yan, K., & Wu, H. (2020). Novel modulation recognizer for frequency-hopping signals based on persistence diagram. IEEE International Symposium on Broadband Multimedia Systems and Broadcasting, BMSB, 2020-October https://doi.org/10.1109/BMSB49480.2020.9379548
