基于Biogeography的SoC测试Wrapper扫描链设计算法  

Wrapper scan chain design algorithm for SoC test based on Biogeography optimization

作　　者：朱爱军[1] 李智[1,2] 许川佩[2] 胡聪[1] 牛军浩[1] 

机构地区：[1]西安电子科技大学机电工程学院,西安710071 [2]桂林电子科技大学电子工程与自动化学院,桂林541004

出　　处：《仪器仪表学报》2012年第12期2774-2780,共7页Chinese Journal of Scientific Instrument

基　　金：国家自然科学基金(60766001)资助项目

摘　　要：基于IP(intellectual property)核的系统级芯片的测试已成为SoC(system on chip)发展中的瓶颈,提出了一种采用BBO(biogeography based optimization)算法的Wrapper扫描链设计方法,使得Wrapper扫描链均衡化,从而达到IP核测试时间最小化的目的。本算法基于群体智能,通过实施迁徙操作和变异操作,实现Wrapper扫描链均衡化设计。本文以ITC'02 Test bench-marks中的典型IP核为实验对象,实验结果表明本算法相比BFD(best fit decrease)等算法,能够进一步缩短Wrapper扫描链,从而缩短IP核测试时间。SoC test based on intellectual property（ IP） core has become a bottleneck of system on chip（SoC） develop- ment. This paper proposes a biogeography based optimization（BBO） algorithm for wrapper scan chain design, which can make wrapper scan chain equalized, so as to achieve the purpose of the minimization of IP core test time. The algorithm is based on swarm intelligence; and can achieve the equalization of wrapper scan chain design through the implementa- tion of migration operation and mutation operation. The typical IP cores in ITC＇02 Test benchmarks were taken as the experiment objects, and experiments were carried out. Experimental results show that the algorithm can shorten wrapper scan chains compared with best fit decrease（ BFD）and other algorithms,thereby reducing the IP core test time.

关 键 词：生物地理学 wrapper扫描链 SOC测试 

分 类 号：TP391.72[自动化与计算机技术—计算机应用技术]