论文部分内容阅读
针对集成电路功能验证中覆盖率收敛较慢的问题,通过分析简单遗传算法(SGA)中精英个体的特征,提出了一种应用于功能验证的精英策略。将本代优秀个体和本代适应度高的历史优秀个体视为精英个体,给予额外交叉机会。基于本文策略的精英遗传算法(EGA)可得到覆盖率广、重复性低的验证向量,缩短功能验证的时间。采用互相关函数的硬件计算单元作为验证模型,在Matlab中模拟功能验证的过程,实验结果表明:与SGA相比,EGA使验证时间缩短了14.8%,功能覆盖率从93%提高到95%,有效地提高了功能验证效率。
Aiming at the problem of slow convergence of IC verification in functional verification, an elite strategy applied to functional verification is proposed by analyzing the characteristics of elite individuals in Simple Genetic Algorithm (SGA). We will regard elitist individuals of this generation as well as elites of outstanding historical significance who are highly adaptable to this generation as elite individuals and give them an extra chance of cross-cutting. Based on the EGA of this paper, we can get the verification vector with wide coverage and low repeatability, and shorten the time of functional verification. The hardware calculation unit with cross-correlation function is used as the verification model to simulate the functional verification process in Matlab. The experimental results show that compared with SGA, EGA can shorten the verification time by 14.8% and the functional coverage from 93% to 95% Effectively improve the efficiency of functional verification.