采用字典词条衍生模式的测试数据压缩

Test Data Compression Using Entry Derivative Mode of Dictionary

为了降低数字集成电路测试成本,压缩预先计算的测试集是一种有效的解决途径。该文根据索引位数远少于字典词条,以及测试数据中存在大量无关位,提出一种采用词条衍生和二级编码的字典压缩方案。该方案引入循环移位操作,确保无关位按序任意移动而不丢失,从而扩大词条衍生性能,减少非词条向量个数。另外,采用规则的两级编码可以减少码字数量和解压电路的复杂度。实验结果表明该文所提方案能够进一步提高测试数据压缩率,减少测试时间。

To lower cost of testing digital integrated circuits,compressing precomputed test set is an effective resolution way.A dictionary compression scheme using entry derivative and two-level coding is proposed based on digits of index far fewer than that of dictionary entry and enormous don＇t-care bits in test data.The introduced cyclic shift operation can arbitrarily shift don＇t-care bits in order without losing them so that derivative performances of entries are expanded and number of non-entry vectors is decreased.In addition,two-level regular coding is adopted to reduce volume of code words and complexity of decompression circuit.The experimental results show that the proposed scheme can farther heighten test data compression ratio and decrease test time.