Abstract: Extracting effective characteristic parameters from the complex radar emitter signals is an effective method to solve the problem of complex radar signal sorting. Because the slice of ambiguity function main ridge of signal can be used to descript the ambiguity energy distribution information of signal, a feature extraction method of the ambiguity function main ridge slice based on graphics contours is proposed. The proposed method can extract two novel features, i.e., cumulated angle and unit cumulative length, which describe the intensity of change and the inherent characteristics of graphics contour, respectively. The experimental results of fuzzy C-means clustering show that the sorting success rate of six types of radar signals, such as normal signal, chirp signal, binary phase shift keying, quaternary phase shift keying, M pseudo-random series, and binary frequency shift keying, is 100% when signal-to-noise ratio (SNR) is no less than 12 dB; and the average success rate of signal sorting is up to 93.5% even if SNR is as low as 0 dB. When SNR changes from 0 dB to 20 dB, the average success rate of signal sorting can reach to 98.9%. The theoretical analysis and the comparative experiment of time consumption for signal sorting show that the proposed method has lower complexity, effectiveness and timeliness.Key

Key words: radarsignal, sliceofambiguityfunctionmainridge, graphiccontour, signalsorting