Work

Automatic Radar Waveform Recognition Using SVM

Year: 2012
Type: article
Abstract: In this paper, a new feature for radar waveform recognition based on the instantaneous frequency is proposed. It is especially utilized for discriminating phase coded signals from other signals. Maximum likelihood estimation (MLE), autocorrelation algorithm, and likelihood ratio test are exploited in the algorithm. In the classification system, support vector machine (SVM) offers an efficient approach to classify linear frequency modulation (LFM) signals, phase coded signals and single frequency signals. Simulation results indicate that the new feature vectors perform effectively over a large range of SNRs. Furthermore, the new classifier achieves a very robust performance that the correct rate is over 90% at SNR of 5 dB, and the ever-increasing rate has been over 97% since SNR of 10dB.
Authors: Hao Gao, Xu Dong Zhang
Institution: Tsinghua University
Language: English
Open Access status: hybrid
OpenAlex ID: W2006493176
FWCI:
Cites: 6
Cited by: 6
Related to: 10