基于DAGSVM的雷达辐射源信号分选与识别

支持向量机具有较好的解决小样本、非线性问题的能力,而DAG算法具有分类精度高的优点。针对现有方法分选与识别准确率不高和对参数变换敏感的问题,在DAGSVM的基础上,提出一种新的雷达辐射源分选与识别方法。首先概述了支持向量机的原理及特点,然后完成了对SVM多分类器的设计,介绍了DAG算法,提出了基于DAGSVM的雷达辐射源信号分选与识别。并通过仿真实验分析了分类器对分选识别结果的影响。实验结果表明,使用DAGSVM这种方法是可行的,该方法具有较强的泛化性能,明显地提高了信号分选识别的准确性。

Sorting and Recognition of Radar Emitter Signal Based on DAGSVM

The support vector machine possesses the ability to solve problems such as small sample,non-linear,and the directed acyclic graphs algorithm has the advantage of high classification accuracy.Because the sorting and recognition accuracy of the common method is not high and is sensitive to the varied parameters,a novel sorting and recognition algorithm for the radar emitter is proposed based on DAGSVM.The principle and features of support vector machine are firstly overviewed in the article.Then,the design of...