KEYWORDS: Contamination, Neural networks, Modulation, Receivers, Matrices, Principal component analysis, Frequency modulation, Digital modulation, Analog modulation, Machine learning, Spectrum analysis, Classification systems
We consider the problem of unsupervised (blind) evaluation and assessment of the quality of data used for deep neural network (DNN) RF signal classification. When neural networks train on noisy or mislabeled data, they often (over-) fit to the noise measurements and faulty labels, which leads to significant performance degradation. Also, DNNs are vulnerable to adversarial attacks, which can considerably reduce their classification performance, with extremely small perturbations of their input. In this paper, we consider a new method based on L1-norm principal-component analysis (PCA) to improve the quality of labeled wireless datasets that are used for training a convolutional neural network (CNN), and a deep residual network (ResNet) for RF signal classification. Experiments with data generated for eleven classes of digital and analog modulated signals show that L1-norm tensor conformity curation of the data identifies and removes from the training dataset inappropriate class instances that appear due to mislabeling and universal black-box adversarial attacks and drastically improves/restores the classification accuracy of the identified deep neural network architectures.
Multi-modal tensor data sets arise with increasing frequency in modern day scientific and engineering applications, for example in biomedical sciences and autonomous engineered systems. Over the past twenty years, tensor-domain data analysis has been attempted primarily in the context of standard (L2-norm) eigenvector decompositions across tensor domains. The algorithms are not joint-tensor-domain optimal and exhibit the familiar sensitivity to faulty/corrupted/missing measurements that characterizes all L2-norm principal-component analysis methods. In this work, we present a robustified method to evaluate the conformity of tensor data entries with respect to the whole accessible data set. Conformity evaluation is based on a continuously refined sequence of calculated L1- norm tensor subspaces. The theoretical developments are illustrated in the context of a multisensor localization application that indicates unprecedented estimation performance and resistance to intermittent disturbances. An electroencephalogram (EEG) data analysis experiment is also presented.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.