基于改进SSD的低截获概率雷达辐射源识别算法  被引量:3

Low Probability of Intercept Radar Emitter Recognition Algorithm Based on Improved SSD

在线阅读下载全文

作  者:赵帅 刘松涛[1] 汪慧阳 ZHAO Shuai;LIU Songtao;WANG Huiyang(Department of Information System, Dalian Naval Academy, Dalian 116018, China)

机构地区:[1]海军大连舰艇学院信息系统系,辽宁大连116018

出  处:《探测与控制学报》2021年第4期103-111,共9页Journal of Detection & Control

基  金:博士后基金资助(2015M572694,2016T90979)。

摘  要:针对低截获概率(LPI)雷达辐射源识别领域中,目标检测类方法实时性不高的问题,提出基于改进SSD的LPI雷达辐射源识别算法。该算法采用精度更高、速度更快的ResNet-50代替VGG-16网络,并对网络进行改进设计;此外,通过构建多径莱斯衰落信道仿真更贴近真实环境的数据集,以提高模型的实用性。仿真实验结果表明,针对7种LPI雷达信号的识别,在信噪比为-4 dB时识别准确率可达到92.9%,与原始SSD模型相比,训练及检测速度分别提高了28.50%和29.74%。Aiming at the low real-time performance of emitter recognition algorithm based on target detection in the field of low probability of intercept(LPI)radars recognition,a radar emitter recognition algorithm based on improved SSD(single shot multi box detector)was proposed.The algorithm used improved ResNet-50 with higher accuracy and faster speed instead of VGG-16 network.In order to improve the practicality of the model,a data set was constructed which was closer to the real environment by building multipath Rician fading channel simulation.The experimental results showed that the recognition accuracy of emitter recognition for seven LPI radar signals could reach 92.9%when the SNR was-4 dB.Compared with the original SSD model,the training and detection speed were improved by 28.50%and 29.74%respectively.

关 键 词:低截获概率雷达 辐射源识别 SSD模型 莱斯衰落信道 

分 类 号:TN973[电子电信—信号与信息处理]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象