Based On Low-resolution Radar Target Identification Method

Target identification is an important development direction in modern radar technology, especially target recognition based on low-resolution radar has important practical significance and value in the current radar system. The methods of target identification based on low-resolution radar are discussed through three basic steps: the radar data preprocessing, the target feature extraction and the classification.First, the process of the radar echo data preprocessing is introduced. The pretreatment is achieved based on tow-FFT, and the disturbance from all kinds of clutter can be restrained, the distance, the spectrum, and other information of a target are obtained.Then, the methods of the target feature extraction are discussed. For the six targets in the paper, a variety of methods for the target feature extraction are used, such as RCS, SVD decomposition, short-time Fourier transform and instantaneous frequency. For each feature extraction way, in order to analyze the stability of the feature under different signal to noise ratio, the signal to noise ratio is reduced through adding Gaussian noise signal to the echo data.Finally, the results of target classification are given. In a brief analysis of the possibility for using single feature for the target identification, the methods of the combination of different features are mainly discussed. According to simulation of the existing experimental data for these targets, the identification results are given with different combination of features under different signal to noise ratio, and under High SNR, the average recognition rate of the targets reaches 85 percent, and the average recognition rate is over 80 percent under lower SNR.