基于 AlexNet 的自适应杂波智能抑制方法

Due to the influence of observation environment and radar platform parameters, the clutter sample data of constructing covariance matrix does not satisfy independent identical distribution, which results in the performance deterioration of the traditional adaptive clutter suppression method. In this paper, an adaptive clutter intelligent suppression method based on AlexNet is developed for stationary radar platforms. Firstly, the sample datasets are established by analyzing the amplitude characteristics of sea clutter. Secondly, by transferring the AlexNet classification model on the ImageNet data-set, and then using the clutter datasets to fine-tune the network parameters. In doing so, sufficient clutter sample data with the independent identical distribution are obtained based on the accurate classification, which improves the performance of the adaptive clutter suppression method. Compared with the existing clutter suppression methods, the proposed method has the advantages in artificial participation, clutter classification, and clutter suppression performances. Finally, the effectiveness of the proposed method is verified by the measured data of CSIR datasets.