突发自适应调制信号的调制识别算法研究

针对突发自适应调制信号中的PSK和QAM调制方式识别问题,本文提出了一种能够识别BPSK、QPSK、8PSK以及16QAM、32QAM、64QAM、128QAM、256QAM八种信号类型的盲识别算法。该算法首先对信号的循环平稳性进行了分析和讨论,给出了利用循环高阶累积量的特征实现信号识别分类的理论依据。然后,提出了三种基于循环累积量的特征分别实现了QAM和PSK类间识别、MPSK类内识别以及方形QAM与十字形QAM的识别。最后通过对MQAM信号的瞬时幅度分布特性的深入研究和分析,提出了一种基于瞬时包络平方的方差的特征实现了QAM的类内识别。该算法选择了二叉树支持向量机作为识别分类器,并设计了一种新的识别流程完成了对上述信号调制方式的识别。该算法无需精确同步,对载波相位具有较好的鲁棒性,并能够对中频信号进行识别。仿真实验表明,该算法能够实现在较低信噪比条件下突发信号的识别。      

适用于超宽带的检测与避让测试平台

文章首先介绍了超宽带技术的特点,分析了目前超宽带工作环境中的干扰问题,引出了缓和干扰的检测与避让机制;接着对欧洲和中国的检测与避让机制的技术要求进行了说明;然后提出了验证检测与避让机制时的面临难点和挑战并给出了解决方案;最后介绍了两种检测与避让测试平台的设计和实现。     

掺砷碲镉汞的光致发光光谱和电学性质

碲镉汞的电学性能和光学性能直接决定探测器性能.对窄禁带掺砷碲镉汞进行了11 ～ 300 K的红外光致发光光谱和变温霍尔测量.对变温光致发光光谱进行了拟合分析,结果表明经通用的两步退火,样品不仅存在砷占碲位(AsTe)、汞空位(VHg)和TeHg-VHg对,还存在碲反位(TeHg).掺杂浓度越大,退火产生的TeHg-VHg对越多.变温霍尔分析进一步证明通用的两步退火后材料中存在TeHg,TeHg留在材料中使迁移率减小.     

Hybrid Perovskite Quantum Dot/Non-Fullerene Molecule Solar Cells with Efficiency Over 15%

Organic-inorganic hybrid film using conjugated materials and quantum dots (QDs) are of great interest for solution-processed optoelectronic devices, including photovoltaics (PVs). However, it is still challenging to fabricate conductive hybrid films to maximize their PV performance. Herein, for the first time, superior PV performance of hybrid solar cells consisting of CsPbI₃ perovskite QDs and Y6 series non-fullerene molecules is demonstrated and further highlights their importance on hybrid device design. In specific, a hybrid active layer is developed using CsPbI₃ QDs and non-fullerene molecules, enabling a type-II energy alignment for efficient charge transfer and extraction. Additionally, the non-fullerene molecules can well passivate the QDs, reducing surface defects and energetic disorder. The champion CsPbI₃ QD/Y6-F hybrid device has a record-high efficiency of 15.05% for QD/organic hybrid PV devices, paving a new way to construct solution-processable hybrid film for efficient optoelectronic devices.     

Scalable 3D Self-Assembly of MXene Films for Flexible Sandwich and Microsized Supercapacitors

The self-assembly of large-area MXene films is the main step to realize their applications in various energy storage devices. However, the scalable self-assembly of flexible thin MXene films with high conductivity as well as excellent mechanical and electrochemical properties is still a challenge. Herein, a synchronous reduction and self-assembly strategy to fabricate flexible MXene films is developed, where MXene films are synchronously reduced and self-assembled on the Zn foil surface. Furthermore, the self-assembly of MXene films can be scaled up by controlling the area of Zn substrates. By adjusting the patterns of Zn substrates, the interdigital MXene patterns can also be obtained via a selectively reducing/assembling process. The resultant MXene films demonstrate high electrical conductivity, large specific surface area, and excellent mechanical properties. Thus they can serve as the electrodes of flexible supercapacitor devices directly. As a proof of concept, flexible sandwich and microsized supercapacitors are designed based on the above MXene film electrodes. Both sandwich and microsized supercapacitors display stable electrochemical performance under various bending states. This study provides a route to achieve large-area MXene-based films or microsized structures for applications in the field of energy storage.     

Improved nanocrystal formation,quantum confinement and carrier transport properties of doped Si quantum dot superlattices for third generation photovoltaics

An all‐Si tandem solar cell has the potential to achieve high conversion efficiency at low cost. However, the selection and synthesis of candidate material remain challenging. In this work, we show that the conventional ‘Si quantum dots (Si QDs) in SiO₂ matrix’ approach can lead to the formation of over‐sized Si nanocrystals especially when doped with phosphorous, making the size‐dependent quantum confinement less effective. Also, our investigation has shown that the high resistivity of this material has become the performance bottleneck of the solar cell. To resolve these matters, we propose a new design based on Si QDs embedded in a SiO₂/Si₃N₄ hybrid matrix. By replacing the SiO₂ tunnel barriers by the Si₃N₄ layers, the new material manages to constrain the growth of doped Si QDs effectively and enhances the apparent band gap, as shown in X‐ray diffraction, Raman, photoluminescence and optical spectroscopic measurements. Besides, electrical characterisation on Si QD/c‐Si heterointerface test structures indicates the new material possesses improved vertical carrier transport properties. Copyright © 2011 John Wiley & Sons, Ltd.     

线性光放大器的宽带自发辐射噪声分析

建立含有垂直光场(VCL)的线性光放大器(LOAs)稳态仿真模型,考虑了宽带噪声特性、端面反射和纵向空间烧孔效应。计算了在不同输入信号功率和垂直腔分布布喇格反射器(DBR)反射率下总的输出ASE噪声功率、噪声系数(NF)、前向和后向传输的ASE噪声光子速率和噪声谱分布的变化。结果表明,与传统光放大器(SOAs)相比,在增益非饱和区,LOA有很好的噪声钳制作用,ASE噪声几乎不随信号功率的改变而变化。在增益饱和区,破坏了VCL的钳制作用,NF有所上升,但上升的速率小于SOA。提高DBR反射率,总的输出ASE噪声功率减小,但NF有所增大。     

激光有源对抗红外成像制导反舰导弹的效能分析

红外成像反舰导弹已经成为水面舰艇的重要威胁之一。随着高功率激光产生技术的迅速发展,激光有源干扰已经成为对抗红外成像反舰导弹一种高效费比的手段。对激光有源干扰的作战效能做了计算和评估,并且对激光有源干扰实现的相关问题做了深入的探讨。     

基于图学习的密集空间网络传输资源调度方法

面对密集空间网络传输资源调度问题中的复杂度以及有效性挑战,本文以图论模型为纽带,将传统数学模型与机器学习方法相结合,提出一种基于图学习的密集空间网络传输资源调度方法.该方法基于图论模型对问题结构的认知将密集空间网络资源调度问题分解,由数学模型与基于图结构的强化学习交替配合完成整个求解过程.实验结果表明,与传统的基于数学模型的资源调度方法相比,所提方法能将资源调度收益提升25.1％,且其训练结果对网络场景变化具有较好的适应性.     

基于径向基神经网络的需求侧响应用户基线负荷计算方法

用户基线负荷是需求侧响应事件中用户响应性能的评估基础。用户基线负荷计算方法应当满足实时监测用户响应性能的要求。本文提出一种基于径向基人工神经网络的用户基线负荷计算方法。该方法考虑非工作日和可能的非正常用电日的影响,选择工作日中的典型日和事件发生前用户的用电情况构成模型的输入向量,训练神经网络。用训练完成的神经网络对假设的系统紧急情况发生时段的负荷进行预测,得到用户基线负荷曲线。实践表明,所提出模型具有一定的预测精度,满足实时性要求。