Composite Interference Signal Feature Extraction And Intelligent Recognition Algorithm Research

In modern electromagnetic environment, the high density, overlapping in time and frequency domain, complex modulation and changeable parameters of signals bring a lot of difficult in electronic warfare. The composited inference signals become a key character in radar jamming and counter jamming field. The traditional ways to extract the radar signal are no longer effective in the complex environment. The methods of feature extraction and recognition of radar composite interference signals need to be figured out promptly. The main issues of this dissertation are mentioned as follow:1. The generation and mathematic model of barrage jamming and deception jamming are analyzed in common use in current electronic warfare environment. After several different ways of composition, the composited jamming signals,which can effectively affect the radar receiving system,are obtained.2. A bunch of common features of composited jamming signal are extracted and the feature distributions of jamming signals are analyzed when JNR varies. By computer simulation,we find out that the recognition result of the common features are not as satisfying as when they are used to recognize single barrage or deception jamming.3. The radar composited jamming signals are abnormal and unstable time sequences. We have extracted the complexity feature , resemblance and several other features of it. We also present the distributions of these features when JNR varies and demonstrate the availability of these features.4. Through several feature optimization and selection strategy, the extracted features are optimized and the key features are selected. Based on these key features,we use neural networks and fuzzy method to recognize radar composite interference. The recognition ratio is above 85%.