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％.     

Event-triggered design for discrete-time nonlinear systems with control constraints

Nonlinear Dynamics - In order to solve the constrained-input problem and reduce the computing resources, a novel event-triggered optimal control method is proposed for a class of discrete-time...     

Statistical Piezotronic Effect in Nanocrystal Bulk by Anisotropic Geometry Control

Utilizing inner-crystal piezoelectric polarization charges to control carrier transport across a metal-semiconductor or semiconductor–semiconductor interface, piezotronic effect has great potential applications in smart micro/nano-electromechanical system (MEMS/NEMS), human-machine interfacing, and nanorobotics. However, current research on piezotronics has mainly focused on systems with only one or rather limited interfaces. Here, the statistical piezotronic effect is reported in ZnO bulk composited of nanoplatelets, of which the strain/stress-induced piezo-potential at the crystals’ interfaces can effectively gate the electrical transport of ZnO bulk. It is a statistical phenomenon of piezotronic modification of large numbers of interfaces, and the crystal orientation of inner ZnO nanoplatelets strongly influence the transport property of ZnO bulk. With optimum preferred orientation of ZnO nanoplatelets, the bulk exhibits an increased conductivity with decreasing stress at a high pressure range of 200–400 MPa, which has not been observed previously in bulk. A maximum sensitivity of 1.149 µS m⁻¹ MPa⁻¹ and a corresponding gauge factor of 467–589 have been achieved. As a statistical phenomenon of many piezotronic interfaces modulation, the proposed statistical piezotronic effect extends the connotation of piezotronics and promotes its practical applications in intelligent sensing.     

计算机网络安全系统设计

分析了现代计算机网络所面临的威胁和各种安全问题,同时提出了在设计一个有关网络系统时应该考虑的一些安全问题,如网络安全策略与技术、网络的访问控制、网络数据加密技术、信息隐藏技术等安全问题。     

草酸钙在硫酸软骨素自组装膜上周期沉淀的研究

利用硫酸软骨素(chondroitin sulfate，CS)在云母基底上通过浇铸法制备的自组装膜为基底，诱导草酸钙在其上的凝集生长。发现当硫酸软骨素的浓度为1．Omg／ml时在该膜体系中可形成规整的周期性草酸钙环状沉淀。这种有序的环状结构可能是耗散结构的一种具体表现形式。利用原子力显微镜(atomic forcemicroscope，AFM)和傅立叶红外光谱仪对这种结构进行了表征，实验结果显示合适浓度下形成的CS膜在一定程度上可以抑制草酸钙的凝集结晶，表明高分子基质与无机离子间强烈的相互作用对无机盐的成核结晶有显著影响，为探讨结石的形成与抑制提供了一定的实验依据。     

16信道0.35μmCMOS/VCSEL光发射模块

研究了垂直腔面发射激光器 (VCSEL)及其列阵器件的光谱特性、调制特性、高频特性及与微电子电路的兼容性 ,将 1× 16的VCSEL与CMOS专用集成电路进行多芯片组装 (MCM ) ,混合集成为 16信道VCSEL光发射功能模块 .测试过程中 ,功能模块的光电特性及其均匀性良好 ,测量的 - 3dB频带宽度大于 2GHz.     

Σ-ΔA/D转换技术及仿真

Σ－ΔＡ／Ｄ转换技术近年来颇受重视，因为基于该技术的Ａ／Ｄ转换器能达到很高的分辨率（１６ｂｉｔ以上），并另具一系列优点。本文对Σ－ΔＡ／Ｄ转换器的基本原理，包括过采样（Ｏｖｅｒｓａｍｐｌｉｎｇ），噪声成形（ＮｏｉｓｅＳｈａｐｉｎｇ），以及数字抽取滤波（ＤｉｇｉｔａｌＤｅｃｉｍａｔｉｏｎＦｉｌｔｅｒｉｎｇ）等内容，均作了叙述。作者还对一个２０ｂｉｔΣ－Δ转换器的算法进行了计算机仿真。该转换器采用了高阶ＭＡＳＨ噪声成形技术，而其数字抽取滤波部分则由梳状滤波器与级联的半带滤波器构成。文章中给出了仿真结果。