有机-无机压电材料TMCM-MnCl3的研究

有机-无机压电材料是一种分子铁电体，具有柔性、结构灵活、易成膜、全液相合成及环保节能等优点，可满足新一代薄膜器件及可穿戴设备的需求。该文以三甲基卤代甲基铵(TMXM, X=F, Cl, Br)为有机部分，MnCl₂为无机部分，通过溶液蒸发法制备了具有钙钛矿分子结构的有机-无机压电材料三甲基氯三氯化锰(TMCM-MnCl₃)，并对其分子结构组成、压电、热学、声学及铁电性进行表征。结果表明，TMCM-MnCl₃的压电常数为106 pC/N，居里温度为130 ℃，声阻抗值约为16.5 MRayl，低于压电陶瓷PZT-4(大于33 MRayl)，具有广阔的应用前景。     

基于特征建模的TDD系统大气波导干扰控制体系研究

大气波导干扰是特定气象条件下发生的时分双工（time-division duplex，TDD）系统内干扰，是TDD移动通信系统大规模组网面临的顽疾。在总结分析大气波导干扰成因和分类等的基础上，对大气波导干扰进行建模和表征，验证了海量干扰源在时域和频域的功率集总特征，并结合大量4G/5G现网实测数据给出了典型条件下内陆波导和海面波导的量化干扰信号传播模型，对于干扰的预测和预防具有重要意义。基于干扰特征，给出了TDD系统预防大气波导干扰的帧结构与组网的4项设计原则，5G现网数据表明干扰控制方案有效，上行干扰下降10 dB以上，相关原则对于6G系统的设计也具有指导意义。     

Eco-friendly and degradable red phosphorus nanoparticles for rapid microbial sterilization under visible light

Pathogens pose a serious challenge to environmental sanitation and a threat to public health.The frequent use of chemicals for sterilization in recent years has not only caused secondary damage to the environment but also increased pathogen resistance to drugs,which further threatens public health.To address this issue,the use of non-chemical antibacterial means has become a new trend for environmental disinfection.In this study,we developed red phosphorus nanoparticles(RPNPs),a safe and degradable photosensitive material with good photocatalytic and photothermal properties.The red phosphorus nanoparticles were prepared using a template method and ultrasonication.Under the irradiation of simulated sunlight for 20 min,the RPNPs exhibited an efficiency of 99.98%in killing Staphylococcus aureus due to their excellent photocatalytic and photothermal abilities.Transmission electron microscopy and ultraviolet–visible spectroscopy revealed that the RPNPs exhibited degradability within eight weeks.Both the RPNPs and their degradation products were nontoxic to fibroblast cells.Therefore,such RPNPs are expected to be used as a new type of low-cost,efficient,degradable,biocompatible,and eco-friendly photosensitive material for environmental disinfection.     

基于1D CNN与XGBoost的恶意代码纹理检测

恶意代码数量已经呈现爆炸式增长,对于恶意代码的检测防护显得尤为重要.近几年,基于深度学习的恶意代码检测方法开始出现,基于此,提出一种新的检测方法,将恶意代码二进制文件转化为十进制数组,并利用一维卷积神经网络(1 Dimention Convolutional Neural Networks,1D CNN)对数组进行分类和识别.针对代码家族之间数量不平衡的现象,该算法选择在分类预测上表现良好的XGBoost,并对Vision Research Lab中的25个不同恶意软件家族的9458个恶意软件样本进行了实验.实验结果表明,所提的方法分类预测精度达到了97％.