基于测试向量相容的编码压缩方法

Compression method based on test vector compatibility

为进一步提高测试数据压缩率,提出一种新的编码压缩方法(FDR-BC).根据测试集中向量间相同位置绝大部分相容的情况,把相容位个数最多的向量划分为一组并合并为一个向量,不相容的位赋值1.用FDR-BC对合并后的向量编码.对分组内向量的个数、不相容位的个数以及不相容位上的值进行编码得到组头编码.解码器结构由输入控制部分和解码部分组成,输入控制部分将编码存储,解码时可以重复利用.解码部分解码的同时移入编码,很大程度上缩短了测试时间.针对ISCAS-89基准电路测试向量集的实验结果表明,用FDR-BC方法进行压缩,相比FDR方法压缩,压缩率平均提升了19.95%.

To improve the compression ratio of test data further,a new method of frequency directed run length based on compatibility （FDR-BC） is proposed.Test vectors that have most compatible bits are grouped and merged together by assigning ＂1＂ to incompatible bits.The merged test vector is encoded with FDR-BC.The numbers of test vectors,incompatible bits,and the values of incompatible bits in the group are encoded to make up the group head code.The decoder architecture consists of input control circuit and decoding circuit.The input control circuit stores the code to reuse.And the decoding circuit can decode and shift the code in simultaneously,which shorten the test time greatly.The experiment we did with the test vector set of the ISCAS-89 benchmark circuit,using the FDR-BC method we proposed,shows that the average compression ratio is increased by 19.95% compared to FDR method.