微波无源器件非线性的实时特征提取与参数化网络建模研究

As the rapidly popularized 5G communication and internet of thing, high density and integrated microwave transceivers emerge quickly, the tight transceiver layout deteriorates the passive nonlinearity issues, its nonlinear spectrum distortion bring huge challenge for high performance microwave communication technology, which already becomes a thorny issue in the modern microwave device research and development. This project takes the nonlinear effect induced harmonic and intermodulation as nonlinear characteristic parameter, firstly develop the real-time differential near field scanning and reference injection-based techniques for locating and quantifying the nonlinear sources in open-space and confined-space respectively. Based on the automatic nonlinear quantization with its big data result, utilizing the neutral network method, a phenomenological nonlinearity model concerns multi-factor coupling effect is built. Finally based on the nonlinear components equivalent, the nonlinear microwave network theory is established. Taking the nonlinearities on the passive devices with certain geometric topology properties, such as filter, power divider, coupler, as case analysis, the established nonlinear microwave network theory is to direct the nonlinearity analysis of complex engineered passive devices. This project will enrich and develop the research foundation of microwave nonlinearities related real-time detection, the parameterized nonlinear network modeling of passive devices and networks. The prospect achievements will be competitive to build international intellectual properties.

随着现代5G通信及物联网技术的快速普及，高密度、高集成的微波收发前端大量涌现。紧凑的收发前端布局加剧了无源非线性效应，其带来的频谱杂散问题给高性能微波通信带来了新挑战，并成为现代微波无源器件的发展瓶颈。本项目以无源器件的谐波及互调杂散信号为非线性表征参数，首先开发差分方向性近场扫描及非线性参考注入方法分别用于开放和封闭空间内非线性源的识别及特征提取；再基于自动化实时的非线性量化过程提取确诊的非线性源大数据，使用神经网络方法建立多参数融合下的无源器件非线性参数化唯象模型；最后结合非线性元件等效探索非线性微波网络分析理论，并对具有一定结构拓扑特征的典型微波无源器件及其网络如滤波器、功分器、耦合器的非线性效应进行案例分析，旨在指导实际复杂无源器件非线性评估。本项目将分别在微波非线性的实时特征提取、无源器件及无源网络的非线性参数化分析建模方面丰富和发展现有国内外研究基础，争夺国际先进知识产权。

随着现代5G通信、物联网与低轨道卫星互联网技术的快速普及，高密度、高集成的微波收发前端大量涌现并成为现代微波器件设计的重要方向。紧凑的收发前端布局显著提升了微波器件的功率/体积比加剧了无源非线性干扰效应，与此同时无源器件的非线性多参数敏感性导致了无源器件的非线性产物在大功率微波通信场景下分析复杂，所带来的频谱杂散问题给高性能微波通信带来了严峻挑战，并成为现代微波无源器件的发展瓶颈。本项目以无源器件的谐波及互调等非线性杂散信号为主要非线性表征参数，分别从检测、建模与对消方面研究了微波无源器件的多参数对非线性产物的影响规律:开发了差分的方向性近场扫描探头实现了对开放空间的辐射信号检测与定位；针对平面微波电路及其表贴元件的非线性干扰识别研发了基于波导加强场的检测方法实现高灵敏度非线性产物检测；借助可控的非线性发生装置实现了封闭空间内非线性源对消研究；使用神经网络优化了非线性干扰的建模过程；针对同轴连接器的非线性产生机理研究了多物理场参数模型与基于DOE方法的多参数拟合模型，均获得了较高的非线性产物预测精度；证明了光照条件对接触结非线性产物的调控关系，提出了接触结电驰豫模型；针对具有二维分布特性与不均匀激励模式的非线性面源干扰产物研究证明了基于互易原理的计算方法，为具有拓扑特征的典型微波器件非线性产物评估提供了灵活高效的评估方法；针对具有多分布非线性源的微波网络的非线性产物计算提出非线性级联模型，证明了待测件电长度和传输反射传播非线性产物关系，进而为判定非线性产物测试有效性提供了参考。综上所述，本项目原定任务顺利完成并额外研究了其他关键问题，获得了大量的数据与模型积累，所研究成果在微波非线性的实时特征提取、无源器件及无源网络的非线性参数化分析建模方面一定程度丰富和发展现有国内外研究基础，争夺国际先进知识产权。