内建自测试多特征混淆模型  

Aliasing probability model with multiple characteristics in BIST

作　　者：郑文荣[1] 王树宗[1] 朱华兵[1] 

机构地区：[1]海军工程大学兵器工程系,武汉430033

出　　处：《海军工程大学学报》2010年第2期74-78,共5页Journal of Naval University of Engineering

摘　　要：特征分析广泛应用于内建自测试体系的响应分析中,通过建立多输入线性移位寄存器的2状态马尔可夫模型和4状态马尔可夫模型,研究了时间无关空间相关及时空均相关下的响应分析器测试序列长度与混淆概率关系,获得混淆的精确数学模型。通过对某控制系统单多特征内建自测试体系混淆结果比较表明:多特征分析可获得更小的混淆概率和更高的测试效率。Characteristic analysis is widely used in the process of response analysis for the architecture of BIST.Models of 2-state Markov process and 4-state Markov process were established for the multiple input signature register（MISR）,which was used for studying the relations between the test length and aliasing error probability for the response analysis under the conditions of the time-independence space-dependence model and time space-dependence model.The accurate expression of aliasing error probability was derived.According to the result of aliasing probability for some control system,the method of multiple characteristics analysis for BIST was presented to achieve the lower aliasing error probability and higher test efficiency.

关 键 词：内建自测试 线性反馈移位寄存器 混淆概率 马尔可夫过程 

分 类 号：TN407[电子电信—微电子学与固体电子学]