行为级DSP算法描述的阵列处理器综合方法研究

本文提出了针对递归DSP算法的高层次系统综合流程,并以脉动(systolic)式处理器阵列结构实现.从DSP算法的FDDL行为级描述开始,经由编译及划分,产生数据相关流图(Data Dependency Graph),然后实现对算法流图的空间映射及时域规划,得到算法的信号流图(Signal Flow Graph),经时序重构,生成脉动阵列,最后实现对处理器单元的数据路径综合及控制器综合,并对处理器单元定位,本文同时提出了各设计阶段的算法策略及优化策略,并给出综合结果。     

武钢四烧工程多目标优化方法烧结试验

采用二次回归正交设计进行烧结试验，获得了七个目标的数学模型，并验证了其可靠性，用多目标计算机选优法对四烧车间的工艺参数进行了选优。     

用于X光激光实验的反射式线聚焦光学系统的设计和分析

基于三维光线追迹的象差计算,对卢瑟福实验室用离轴球面反射线聚焦系统作了分析。然后,首次提出了用于X光激光实验的两类新的反射式线聚焦构型,即反射式光楔列阵和光锥列阵系统,并作了计算。最后,对所得结果作了比较和讨论。     

异步时序型PAL解析方法的研究及解析结果分析

本文就异步时序现ＰＡＬ的解析，提出了一种状态图遍历算法。     

基于小波基优化和动态比特分配的数字音频压缩编码方案

提出了一种新颖的基于自适应小波基优化选择和心理声学模型相结合的数字音频信号的透明质量编码方法，保证固定失真水平上使每帧信号的变换系数的动态分配的比特数最少，并且利用动态码本的方法来消除音频信号的统计冗余，进一步压缩比特率，对于抽样率为４４．１ｋＨｚ每样值用１６比特线性码表示的光盘单声道音乐信号可以压缩到６４ｋＢＰＳ左右。     

大孔距小排距的孔网参数在中深孔爆破中的应用

天铁集团的石矿厂多年来爆破效果不理想，粉矿率高，通过对爆破孔网参数的优化，采用大孔距、小排距的孔网参数在中深孔爆破中取得了较好的效果。     

紧凑型FLC光纤PDL模块

描述了紧凑型FLC（铁电液晶）光纤PDL（光子延迟线）模块的结构和运转特性；分析了插入损失；提出多信道运转时依赖于波长的设计方案。     

C语言中浮点数输入出现异常的研究

在TC2、0的编译环境下，发现利用“scanf0”函数给浮点型二维数组和结构体数组赋值时，没有语法错误，但程序不能正常运行。通过实例分析了产生这种问题的原因并提出了解决方法。     

A Novel UWB Pulse Shape Modulation System

In this paper novel modified Hermite polynomial functions for use in impulse radio (ultra-wideband) communications are proposed. With these functions pulse shapes which are orthogonal and have nearly constant pulse width regardless of the pulse order are generated. These properties hold under the effects of differentiation. An M-ary communication system is constructed using these pulse shapes. A Matlab model for generating the pulses is designed and the effect of timing jitter on the performance of the system is investigated by computer simulation.     

Spreading Codes Generator for Wireless CDMA Networks

The prime characteristic of spread spectrum modulated signals is that their bandwidth is greater than the information rate. In this way a redundancy is introduced that allows the severe levels of inteference inherent in the transmission of digital information over radio and satellite links to be overcome. Current spread spectrum applications are primarily in military communications; nevertheless, there is growing interest in this technique for third generation mobile radio networks (UMTS, FPLTS, etc.) with open discussion regarding the practicality of using a multiple access system based on spread spectrum techniques (CDMA). However, in order to support as many users in the same bandwidth as other multiple access techniques such as TDMA or FDMA, it is important how to generate large families of sequences that present low cross-correlation. The aim of this paper is to describe a spreading codes generator that can produce a large number of PN sequences with good properties of auto- and cross-correlation. Moreover, the codes generated shows high unpredictability and good statistical behaviour. This also allows the implementation of some features that are common on military networks such as message privacy (increasingly important in commercial networks) without additional cost. The structure presented shows itself to be advantageous for high speed generation of codes at a low cost, low power consumption (allowing longer life for batteries), small size and simplicity of implementation, essential ingredients for commercial equipment. Another attractive feature is its structural parallelism, useful in VLSI implementations. All of these features render it potentially suitable for the implementation of channel bandwidth sharing systems in future wireless personal communications networks.