Abstract
Automatic modulation classification (AMC) is playing an increasingly important role in spectrum monitoring and cognitive radio. As communication and electronic technologies develop, the electromagnetic environment becomes increasingly complex. The high background noise level and large dynamic input have become the key problems for AMC. This paper proposes a feature fusion scheme based on deep learning, which attempts to fuse features from different domains of the input signal to obtain a more stable and efficient representation of the signal modulation types. We consider the complementarity among features that can be used to suppress the influence of the background noise interference and large dynamic range of the received (intercepted) signals. Specifically, the time-series signals are transformed into the frequency domain by Fast Fourier transform (FFT) and Welch power spectrum analysis, followed by the convolutional neural network (CNN) and stacked auto-encoder (SAE), respectively, for detailed and stable frequency-domain feature representations. Considering the complementary information in the time domain, the instantaneous amplitude (phase) statistics and higher-order cumulants (HOC) are extracted as the statistical features for fusion. Based on the fused features, a probabilistic neural network (PNN) is designed for automatic modulation classification. The simulation results demonstrate the superior performance of the proposed method. It is worth noting that the classification accuracy can reach 99.8% in the case when signal-to-noise ratio (SNR) is 0 dB.
Highlights
Automatic modulation classification is a technology for the automatic classification of signal modulation types [1], which is widely used in interference identification, spectrum sensing, electronic countermeasures and other fields
We generated 84,000 signals for all the modulation types, with signal-to-noise ratio (SNR) continuously randomly chosen from −20 to 20 dB
Considering the background noise −100 dBHz, bandwidth 5 MHz and the sampling rate 50 MHz, the input SNR from −20 to 20 dB is equivalent to the input signal power of from −70 to −30 dBm approximately
Summary
Automatic modulation classification is a technology for the automatic classification of signal modulation types [1], which is widely used in interference identification, spectrum sensing, electronic countermeasures and other fields. Feature-based methods consisting of feature extraction and classifier design, have been widely applied in the AMC field to recognize more modulation types with lower algorithm complexity (see [5]). In [16], an AMC algorithm using CNN to extract constellation features of digital communication signals is studied, which combines image classification and deep learning. An AMC scheme based on deep learning and feature fusion is proposed, which can achieve high classification accuracy in a low SNR environment, and within a large range of SNR. In [18], an image fusion algorithm is proposed for modulation classification, which considers the images from different time-frequency methods.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have