摘要
测试数据量的快速增长成为导致测试时间增加、制造成本提高的一个主要因素。为减少测试数据量,提出一种快速查找最佳有理渐近分数的编码压缩方法。不直接存储游程数据,而是将游程数据转换成浮点数,快速查找浮点数对应的最佳有理渐近分数,最终以整数分子和整数分母的形式存储游程出现的规律。该方法相容于传统的编码方法,压缩和解压协议简单,压缩效果好,硬件开销小。仿真实验结果证明了该方法的有效性和稳定性,与国内外同类方法相比具有一定优势。
Rapid growth of test data volume becomes a major factor for test time and manufacturing cost increasing. To reduce test data volume, a code-based compression scheme with fast search best rational approximate fraction is presented. The run-length data is converted into floating point numbers; and the equal best rational approximate fractions of floating point numbers is searched quickly; the appearing law of run-length data in the form of integer numerator and integer denominator is stored instead of storing run-length data directly. This scheme is compatible with traditional code-based methods. It also has simple compression and decompression protocol, good compression effect and small hardware overhead. Simulation results demonstrate the effectiveness and stability of the scheme. Compared with the similar method at home and abroad, the scheme also has certain advantages.
作者
吴海峰
詹文法
程一飞
Wu Haifeng;Zhan Wenfa;Cheng Yifei(School of Computer and Information, Anqing Normal University, Anqing 246011, China;Department of Science Research, Anqing Normal University, Anqing 246011, China)
出处
《系统仿真学报》
CAS
CSCD
北大核心
2018年第6期2384-2389,共6页
Journal of System Simulation
基金
国家自然科学基金(61540011
6130646)
安徽省学术技术带头人后备人选(GXBJZD2016075
2015H053)
安徽高校自然科学研究重点项目(KJ2016A427
KJ2017A351)
关键词
测试数据压缩
最佳有理渐近分数
游程长度
浮点数
test data compression
best rational approximate fraction
run-length
floating point number
