基于卷积神经网络的信号码速率估计方法  

Symbol Rate Estimation Based on Convolutional Neural Network

作　　者：杨欣 叶云霞 陈媛 YANG Xin;YE Yun-Xia;CHEN Yuan(The 36^(th)Research Institute of CETC,Jaxing 314033,China;National Key Laboratory of Electromagnetic Space Security,Jaxing 314033,China)

机构地区：[1]中国电子科技集团公司第三十六研究所,浙江嘉兴314033 [2]电磁空间安全全国重点实验室,浙江嘉兴314033

出　　处：《中国电子科学研究院学报》2024年第7期634-638,共5页Journal of China Academy of Electronics and Information Technology

摘　　要：在电子侦察中目标对象的码速率是一个重要的信号参数,该参数的获取是进一步对信号进行解调解译的前提条件,因而信号码速率的估计具有重要的现实意义。近年来,随着人工智能技术的飞速发展,深度学习在信号处理领域已得到了较为广泛的应用。本文提出了一种基于卷积神经网络的信号码速率估计方法,直接将目标信号的采样数据作为网络输入即可得到对应的估计结果。该方法无需信号调制样式先验已知,且对不同调制样式信号的估计流程一致,因此简化了算法流程,最后通过仿真对算法的有效性进行了验证。仿真结果表明,该方法可以实现信号码速率的准确估计,且对短猝发信号和低信噪比场景具有良好的适应性。In electronic reconnaissance,the symbol rate of target signal is an important parameter,and acquiring this parameter is a prerequisite for further demodulation and interpretation of the signal.Therefore,symbol rate estimation is of great practical significance.In recent years,with the rapid development of artificial intelligence technology,deep learning has been widely applied in the field of signal processing This article proposes a symbol rate estimation method based on convolutional neural networks,which directly takes the sampled data of target signal as network input to obtain the corresponding estimation results.This method does not require prior knowledge of signal modulation style,and the estimation process for signals with different modulation styles is consistent.So the algorithm process has been simplified.The effectiveness of the algorithm is verified by simulation.The simulation results show that this method can estimate the signal symbol rate accurately and has good adaptability to short burst signals and low signal-to-noise ratio scenarios.

关 键 词：深度学习 卷积神经网络 码速率估计 

分 类 号：TN911.7[电子电信—通信与信息系统]