S Mode Signature Analysis And Identification Method

The recognition of the property of Identification Friend or Foe is the extraordinary kernel and pivotal point. Not only in the daily management of aviation, but also in the aerial warfare, Mode S signal is the key-point to realize the normal air traffic control, the liaison in wartime and recognition of identity. It is significant for national defense to study recognition of Mode S signal. During the research of signal recognition method, feature analysis and extraction is a pivotal technique in Mode recognition which influences the precision of signal's recognition directly. The dissertation started with signal format of Mode S, based on feature analysis and feature extraction, the recognition methods of Mode S signal are deeply and widely studied in this dissertation. The recognition of Mode S signal is realized by the simulation result.This dissertation primarily investigates the research of feature analysis and recognition method. The main work and achievements of the dissertation can be summarized as follows:1. The produce mechanism of Mode S signal and the characteristic of Mode S base band signal's data framework are analyzed. Specific for Mode S interrogating signal and Mode S request signal, the signal framework, operational principle of Mode S system, operating frequency and types of modulation.2. Specific for Mode S signal, the feature parameters in signal's Time-Domain (such as instantaneous amplitude, instantaneous phase and the standard deviation of the absolute value of the non-linear component of the instantaneous phase), Frequency-Domain (such as the value of the normalized centered instantaneous amplitude) and Time-Frequency Transformation-Domain (Short Time Fourier Transform, Wigner-Ville time-frequency distribution and Wavelet Transform) are analyzed and extracted, then realization the recognition of Mode S signal with the feature parameters which have significance for signal classification.3. Based on the extracted feature parameters which have the classification importance, the statistic feature parameters recognition method, the classification method based on continuous wavelet transform, the fuzzy recognition method based on spectral characteristics are used to realize the recognition of Mode S signal, the simulation results each recognition method's performance in different SNR. Specific for Mode S interrogation signal, a novel algorithm is provided, which can effectively estimate the symbol rate of signal using the Wavelet Transform and the spectrum analysis, the simulation results this method have good performance.4. From the analysis of 2-order Moment to 4-order Moment and Higher-Order cumulant, 3 parameters which have high differentiated performance and small computational quantity are extracted, analyzing using decision tree simultaneity, the dissertation suggests that the Classifier of hierarchical neural network based on High-order Moment and Higher-Order Cumulant, constructing hierarchical neural network to realize automatic identification which has fast recognition speed and high precision in low SNR.