FDR-COMPRESS

  一种改进的方案，它针对硬故障集的特点利用指示频率连续长度码(FDR码)进行二次编码压缩。该方法充分利用测试集中无关位较多而测试模式之间变化位较少的特征，应用贪心算法对预先计算的测试集进行了重新排序，尽可能的使相邻的向量间具有最少的不同位，排列成一个最佳的差分向量序列，然后应用FDR码对差分向量序列进行编码,最后对压缩结果重新分块，按照一次压缩的流程进行二次压缩。在解压时，使用片上解压电路解压，能高效地还原出原测试集。最后通过对Soc部分标准电路硬故障集的测试结果显示，本方案具有较好的压缩率。(a new improved method was proposed in the paper,the method using Frequency-Directed Run-length(FDR codes)coding for the twice aimed at hard-fault. It fully takes advantage of the characters that there are many don’t-care bits in the test set but a few different bits between the test patterns, Then it applies the greed calculate way to sort the precomputed test sets and make the close together vector have the least different sites possibly. It uses the FDR codes to coding the vectors,then blocks the results and uses the FDR codes to compress for the second time. Decompression used of on-chip circuit extract unpack,it will be deoxidize the original test efficient.using the hard-fault to compress and the experimental results prove that the improved coding method of this paper has the better compression ratios has been proved at last. )