对称M矩阵的并行不完全分解预条件

J.H.Yun提出了一种新的计算块三对角M矩阵预条件的算法,这种方法具有天然的并行性,解决了ILU分解不易并行化的缺点,能有效节约计算时间。以对称M矩阵作为例子,将以上方法推广到一般的对称M矩阵,使得在构造这一类矩阵的不完全分解预条件方法时,能够更加快速有效。关于预条件子的定理及其证明将会被给出。最后,数值实验将会被用来验证我们的定理结论。     

双曲空间运动群Lie子群作用的轨道

双曲空H~3在Weierstrass坐标下的运动群就是Lorentz群G,最近,我们确定了G的各类Lie子群,本文则集中讨论G的2参数Lie予群H在H~3上作用的轨道,主要结果如下:定理1 Lorentz群G的2参数Lie子群H在双曲空间H~3上作用的轨道,或者是圆柱面,或者是非正常曲率的全脐点曲面,反之亦然.定理2 Lorentz群G的每个2参数Lie子群H在双曲空间H~3上作用所得的不同轨道,彼此测地平行.     

AMS-02地面传输系统

为了研究在高能物理试验中使用超长TCP连接进行数据传输的最佳效率,分别在IPv4和IPv6网络协议上进行了一系列的试验寻求优化方法.介绍了AMS-02项目中东南大学地面传输系统的总体结构,以及在CERNET/CERNET2和欧洲粒子实验室之间进行的数据传输试验,试验主要研究了并行流数目和TCP缓冲大小对传输效率的影响.结果表明:在目前CERNET的环境下,找到最佳的并行流数目是提高传输效率的主要方法;TCP缓冲调节在目前的条件下效果不明显,但是当可用带宽显著提高后会有明显的效果.     

GPU并行计算的CUDA架构浅析

本文阐述了GPU并行运算的一种主流架构——CUDA架构,包括CUDA编程模型、程序的运行模式、线程架构、存储器结构、指令结构等。     

基于区域分解的共轭梯度法的有限元结构分析并行计算

1 Introduction Large and complex engineering problems often needmuch computation ti me and storage to run on ordinaryuniprocessor computers . Even if they can be solved ,powerful computation capability is required to obtainaccurate and reliable results within reasonable ti me .Parallel computing can meet requirements of high per-formance computing[1].Various types of parallel hard-ware architectures have been developed and parallelalgorithms adapted to these hardware architecturesproposed[2 ,3…     

Parallel computing for finite element structural analysis using conjugategradient method based on domain decomposition

1 Introduction Large and complex engineering problems often needmuch computation ti me and storage to run on ordinaryuniprocessor computers . Even if they can be solved ,powerful computation capability is required to obtainaccurate and reliable results within reasonable ti me .Parallel computing can meet requirements of high per-formance computing[1].Various types of parallel hard-ware architectures have been developed and parallelalgorithms adapted to these hardware architecturesproposed[2 ,3…     

基于思政建设的专业课改革探讨

土木工程本科专业教育作为城乡建设事业的重要基础,需要和当代思想政治教育相结合,二者并行建设发展。本文在综合分析课程建设的主要内容和涵盖方式的基础上,将思政教育目标有机融入教学大纲,完善教案,改革教学方法,总结经典案例,在此过程中进行教学团队建设,进一步提高教学水平和人才培养质量。     

PARALLEL IMPLEMENTATIONS OF NUMERICAL SIMULATION OF WIND FLOW AROUND BUILDINGS

ＩＮＴＲＯＤＵＣＴＩＯＮＷｉｔｈｔｈｅｄｅｖｅｌｏｐｍｅｎｔｏｆｃｏｍｐｕｔｅｒｔｅｃｈｎｏｌｏｇｙａｎｄｎｕｍｅｒｉｃａｌｍｅｔｈｏｄｓ,ｎｕｍｅｒｉｃａｌｓｉｍｕｌａｔｉｏｎｔｅｃｈｎｉｑｕｅｓｈａｖｅｂｅｃｏｍｅｎｅｗｅｆｆｅｃｔｉｖｅａｐｐｒｏａｃｈｅｓｆｏｒｔｈｅｐｒｅｄｉｃｔｉｏｎｏｆｗｉｎｄｖｅｌｏｃｉｔｉｅｓａｒｏｕｎｄｂｕｉｌｄ ｉｎｇｓａｎｄｗｉｎｄ ｉｎｄｕｃｅｄｐｒｅｓｓｕｒｅｓｏｎｔｈｅｂｕｉｌｄｉｎｇｅｎｖｅｌｏｐｅ (Ｍｕｒａｋａｍｉｅｔａｌ.,1 988;Ｂａｓｋａｒａｎｅｔ…     

关于平行研究

论述比较文学又一研究方法——平行研究的概念、方式、范围、内容、种类和研究的可比性及文学性原则。     

Interference cancellation technique under imperfect synchronization in cellular systems

In this paper, an asynchronous cooperative cellular system applied with space-time block coding(STBC)is investigated. A signal detector is proposed based on parallel interference cancellation(PIC), to cancel the inter-symbol interference(ISI)caused by the imperfect synchronization. Simulation results show that the proposed PIC detector can effectively suppress the ISI, but there is still a comparatively high error floor, due to the co-channel interference(CCI)of the cellular system.