MIMO系统中的新型K-best检测算法(英文)

针对K-best检测算法易将最优路径舍去的特点和K-best检测算法搜索星座图中所有点的特点,提出一种性能改进型K-best检测算法和几种降低复杂度K-best检测算法.性能改进型K-best检测算法在进行QR分解之前对信道矩阵进行最小均方误差(MMSE)滤波,能有效减小最优路径被舍弃的概率,提高算法性能;降低复杂度K-best检测算法采用类似球形译码检测的方法减少搜索星座图中点的个数.仿真结果显示,性能改进型K-best检测算法比基于排序QR分解(SQRD)的K-best检测算法有1dB的性能增益.降低复杂度K-best检测算法在K=4时有性能损失;当K=8时,降低复杂度K-best检测算法和原K-best检测算法有同样的性能,同时前者比后者需要更少的计算量.     

软判决Viterbi译码和序列译码在FPGA上的实现

介绍了3位软判决Viterbi译码器和序列译码器的FPGA实现，其中选用了同样的码率1／2和约束长度7，在FPGA实现的基础上，对Viterbi译码器和序列译码器的译码性能进行了测试和比较。结果表明，虽然序列译码器出现很小的译码增益损失，但其译码速度与Viterbi译码器不相上下，所消耗的硬件资源大大低于Viterbi译码器，当然这是以增加译码延时为代价的。     

言语交际容错阈与语言文字规范化

言语交际中存在编码语义所指错误而解码无误的现象,是因为编码错误在容错阈之内;容错阈的大小既与解码者的语言素质高与低有关,也与编码错误的显与隐、广与狭有关;当一种编码错误非常广泛,已进入"全民容错阈"时,应该就是被语言文字规范化"容错"的对象。     

大学英语写作教学中的文化解码

语言是文化的载体。对语言的理解输入过程可以称之为对文化的解码过程。在大学英语写作过程中,对语言解码的效果直接影响到语言的输出。文中从文化解码的角度出发。分析学生在写作过程中所面临的文化解码难题,并提出提高大学生英语语言编码技巧的对策等。     

语块加工与口译质量的相关性研究

语块以其结构稳定性和功能多样性,在语言信息的编码和解码过程中占有明显的优势。通过话语复述作业、口译作业中语块介入的效能和产出质量,证明语块在英汉、汉英口译中的优势。     

数字化变电站GOOSE通信机制及报文实例分析

GOOSE是IEC 61850中定义的一种通用变电站事件模型类,对于实现设备间的互操作具有重要的意义。介绍了基于IEC 61850标准的GOOSE服务的总体机制,通过对涉及的关键概念进行阐述,以江西500k V马回岭数字化变电站的实际情况为例,对日常运行中检修维护人员需关注的具体问题以抓获报文的方式对其进行分析。并结合调阅监控后台读取SCD配置文件中的模型及虚拟二次回路来检验排故,对继电保护人员在维护、定检工作及故障分析中具有实际指导意义。     

"Julia曲线"与分形图像压缩编码

将 Julia曲线"按正方形形状以多种方式进行量化,并将量化的 Julia曲线 "用于分形图像压缩编码,改变了分形图像压缩编码以变化的压缩编码字典进行编码的缺点。此外,还建立了一个小型的常用字典,用以加速分形图像的压缩编码。实验结果表明, Julia曲线 "能很好地拼贴所要编码的图像,并具有分形图像的解码优点。     

A fast BER evaluation method for LDGM codes

Low-density generator matrix (LDGM) codes have recently drawn researchers’ attention thanks to their satisfying performance while requiring only moderate encoding/decoding complexities as well as to their applicability to network codes. In this paper, we aim to propose a fast simulation method useful to investigate the performance of LDGM code. Supported by the confidence interval analysis, the presented method is, for example, 10⁸ times quicker than the Monte-Carlo computer simulation for bit-error-rate (BER) in 10⁻¹⁰ region.     

Early prediction of individual growth in reading

Seventy four boys with reading disabilities and amatched control group of normal readers were followedduring their compulsory schooling (to grade 9). Worddecoding performance was assessed with the Wordchainstest on three occasions. For each student a lineargrowth function was estimated. The slope parameter wasused as the dependent variable in a multipleregression analysis with a large number of explanatoryfactors from grades 2 and 3 including phonologicalskills, reading and spelling, non-verbal intelligence,motor skills, eye movements, teacher ratings of schoolbehaviour and parents' education. The proportion ofvariance in growth explained for the reading disabledboys amounted to 25%, the most powerful predictorsbeing early reading performance and non-verbalintelligence. More surprisingly, neither teacherratings nor family background gave useful predictiveinformation. For the control group none of theindependent variables explained the variance of growthin reading.     

The transient role of explicit phonological recoding for reading acquisition

This study explores early stages of reading acquisition, specifically therelation of phoneme blending and letter recoding to individual differencesin word decoding. The hypothesis was that facility in letter recoding andaccuracy of phoneme blending are substantial components of word decodingin beginning readers but not for skilled reading and that reliance onphoneme-sized decoding of words would dissipate as reading proficiencyincreased. In four studies we examined the ability to recode letters, blendphonemes and decode words in four groups of Dutch children (early Grade1, N = 130, older Grade 1, N = 81, Grade 3, N = 54 and a group of children witha reading lag, N = 356). Phoneme blending was only related to the readingability of beginning Grade 1 children. By the end of Grade 1 ability to blendphonemes did not differentiate reading capacity, nor for older children inGrade 3 and reading disabled children. Letter recoding was related to worddecoding in all four studies, although the strength of that relation diddwindle as reading skill level increased. The results of the study appearconsistent with self teaching hypothesis and other theories that imply atransient role of explicit phonological recoding in word identification.