PROJECT TITLE :

Robust Automated VHF Modulation Recognition Based on Deep Convolutional Neural Networks - 2018

ABSTRACT:

This letter proposes a completely unique modulation recognition algorithm for terribly high frequency (VHF) radio signals, which is predicated on antinoise processing and deep sparse-filtering convolutional neural network (AN-SF-CNN). First, the cyclic spectra of modulated signals are calculated, and then, low-rank representation is performed on cyclic spectra to scale back disturbances existed in VHF radio signals. When that, before fine tuning the CNN, we tend to propose a sparse-filtering criterion to unsupervised pretrain the network layer-by-layer, which improves generalization effectively. Many experiments are taken on seven types of modulated signals, and the simulation results show that, compared with the ancient ways and some renowned deep learning methods, the proposed methodology can achieve higher or equivalent classification accuracy, and presents robustness against noises.