基于射频基带联合匹配滤波的雷达与通信频谱共享研究

Spectrum sharing between radar and communication systems can solve military and civilian departments' spectrum shortage problems, but it needs to face the challenge of strong in-band interference. In-band interference can saturate analog RF front-end receiver, no matter what kind of digital baseband signal processing algorithms are used. So it not only limits anti-jamming performance of the receiver, but also limits the number of users for spectrum sharing. Therefore, it is necessary to study RF analog matched filter technology based on the baseband matched filter theory. This project focus on radar and communication systems, with efficient sharing of spectrum resources as the core. Matched filter principles and combined RF and baseband simulations are used as basis for our project. We will systematically focus on RF and baseband matched filter mechanisms, scenarios and performance bounds of joint matching, hardware-in-the-loop(HIP) simulation and test methods. And the high performance dispersion delay mechanism which can realize RF matched filter will be revealed. Machine learning based high speed spectrum shape matching method will also be evaluated. We will also prove the information theory boundary used for communication and radar spectrum "isolation". The project is expected to reveal the realization mechanism of dispersion delay matched filtering technology, establish radar and communication signals' template library, build spectrum sharing HIP platform. The longstanding limitation of performance degradation due to high in-band interference of radars, wireless communications, spectrum sensing and electronic warfare systems will also be solved.

雷达与通信的频谱共享会解决军民用频谱资源紧缺的问题，但需要面对带内强干扰的挑战。带内干扰会使得接收机的模拟射频前端饱和，无论在基带的数字部分采用什么信号处理算法都于事无补，这就限制了接收机的抗干扰性，同时也限制了频谱共享的用户数量。因此，在基带匹配滤波基础上，需要研究射频模拟匹配滤波技术。本项目以雷达与通信系统为研究对象，以频谱资源高效共享为核心，以匹配滤波原理和射频基带联合建模为基础，系统研究射频和基带匹配滤波的机理、联合匹配滤波的场景和性能上限、半实物仿真试验测试方法，以及高性能色散延迟器机理、基于机器学习的高速频谱形状匹配搜索方法、用于通信与雷达频谱"隔离"的信息论边界等关键科学问题。项目预期将揭示色散延迟技术实现模拟匹配滤波的机理，建立雷达和通信信号模板库，搭建频谱共享半实物仿真平台，也对解决长期困扰雷达、无线通信、频谱感知和电子对抗等强干扰下的性能下降问题有一定意义。

本项目面向强干扰下的雷达与通信的频谱共享问题，通过射频和基带联合的匹配滤波方法达到干扰抑制，实现高效的频谱共享。项目主要思路是：通过智能电磁信号感知，实现强干扰信号的识别，探索基于机器学习和深度神经网络的信号识别方法；进行模拟信号处理以完成干扰信号的射频抑制，再使用基带匹配器滤波进一步匹配出感兴趣的发射信号；构建了雷达和通信波形库，搭建软件无线电半实物仿真平台，在外场和暗室进行算法性能测试验证。本项目执行期为2019年01月到2021年12月。本项目的预期成果是发表学术论文5篇，发明专利3项，搭建半实物仿真系统一套，完成结题报告一份。在执行期内，发表SCI检索论文3篇、EI期刊论文1篇、中文核心期刊论文1篇；项目负责人申请/授权专利5/1项；搭建了多通道相干雷达通信一体化软件无线电半实物仿真平台一套，在外场和暗室对脉冲与5G的干扰共存进行了验证。基于项目成果转化来的“干扰信号识别”科教融合案例，获得了2021年首届全国高校教师教学创新大赛一等奖、教学活动创新奖、2021年第一届全国高校电子信息类专业课程实验教学案例设计竞赛全国特等奖。2020年12月，项目负责人基于该项目的延伸获得了军委科技委国防科技创新特区基金支持。项目组圆满完成了项目任务。