基于神经网络的雷达抗欺骗干扰方法  

Method against Radar’s Deceptive Jaming Based on Neural Network

作　　者：程高峰[1] 蒋晓瑜[1] 陈昕[1] 李强[1] 

机构地区：[1]装装甲兵工程学院控制工程系,北京100072

出　　处：《装甲兵工程学院学报》2006年第4期74-77,共4页Journal of Academy of Armored Force Engineering

摘　　要：对雷达抗应答式欺骗干扰的特征提取方法进行了改进，统计方法将均值与方差特征相结合定义特征因子，神经网络方法用Kohonen网络进行特征提取。仿真结果表明，2种方法都具有较好的抗应答式欺骗干扰性能，而神经网络方法性能更为优越。Based on the method of radar against answering deception-jamming, this paper advances two new methods of feature extraction. The statistical method defines a feature factor united the mean and variance feature, the NN method extracts the feature with Kohonen network. Results of simulation show that these two methods perform well against deceptive jamming, and the NN method is better in training and detecting outcome.

关 键 词：应答式欺骗干扰 特征提取 神经网络 

分 类 号：TN973.3[电子电信—信号与信息处理]