基于粗糙集的决策树雷达辐射源识别方法

摘要

On the optimization of identifying radar emitter, it is difficult to identify radar emitter signals due to the impact of environment and interference. To solve such problems, a new method is proposed to establish decision tree to recognize radar emitter signals in this paper. Radar emitter signals are processed by Rough Sets first. After discretization and attribute reduction, importance and relative quantity of information of each attribute are obtained. The decision tree can be established based on the relative quantity of information and recognize radar emitter signals. The method introduced in this paper improves the traditional decision tree arithmetic and simplifies the decision tree structure. The simulation results verify the effectiveness and reliability of the method. It gives a basis for eliminating the noise and recognizing results correctly.%关于优化识别雷达辐射源问题,雷达辐射源信号受到外界环境影响和各种干扰,造成识别困难.为解决上述问题,提出了一种新的决策树雷达辐射源识别方法.首先,应用粗糙集理论对雷达辐射源信号进行分析.通过离散化、属性约简,得到分类结果对于各条件属性的重要度.根据各条件属性重要度来确定各属性信息量的相对大小,建立决策树.该方法简化了决策树的结构,并改进了传统的决策树的建立方法.选取雷达信号的主要参数特征进行了仿真.仿真结果验证了改进方法的有效性与可靠性.为消除噪声,正确识别结果提供了依据.