Classification of correlation signatures of spread spectrum signals using neural networks

Abstract

The authors discuss the application of artificial neural networks (ANNs) to the classification of spread spectrum signals based on signal type or spreading technique. Radial basis function networks (RBFNs) and back-propagation networks (BPNs) were used to classify the correlation signatures of the signals. Correlation signatures of four types or classes were obtained from United States Army Harry Diamond Laboratories: direct sequence (DS), linearly stepped frequency hopped (LSFH), randomly driven frequency hopped (RDFH), and a hybrid of direct sequence and randomly driven frequency hopped (HYB). RBFNs and BPNs trained directly on two classes (DS and LSFH) and four classes (DS, LSFH, RDFH, and HYB) of correlation signatures. Classification accuracies ranged from 79% to 92% for the two-class problem and from 70% to 76% for the four-class problem. The RBFNs consistently produced classification accuracies from 5% to 10% higher than accuracies produced by the BPNs. The RBFNs produced this classification advantage in significantly less training for all cases.<>