針對開放式線段缺陷的診斷性自動測試向量產生器與其診斷流程設計

Abstract

當電路發生開放式線段缺陷，實體電路上的缺陷周圍的其他邏輯閘造成的偶和電容及拜占庭效應對缺陷影響到的輸出會隨著缺陷的位置改變，許多先前的研究針對開放式缺陷的測試和檢驗。這篇論文提供了一個利用兩個階段產生診斷性測試向量的方法：首先使用布林可滿足性引擎在單一缺陷的假設下，針對發生缺陷影響到的邏輯閘自動產生向量。另外搭配上兩階段的診斷方法，分別使用一個修正的字典(dictionary)診斷法，去減少可能發生錯誤開放式缺陷的候選線段數量，接著在利用注入和評估(inject-and-evaluate)診斷法使得結果能夠達到很高的準確性。實驗執行在ISCAS85的電路上，結果顯示在檢測方面最後的解析度(即對應的候選線段數量)可以達到幾乎接近1的準確性。

As an open defect occurs on one segment in the circuit, the coupling capacitances from the neighboring nets and the Byzantine effect based on the physical layout and cell library can result in different faulty behaviors. Many previous researches focus on the test and diagnosis issue for open defects. However, the diagnosability of test patterns is not well addressed. Therefore, in this paper, we propose a high-resolution algorithm which consists of two stages: (1) a diagnostic ATPG and (2) its diagnosis flow. A Boolean Satisfiability solver is incorporated to generate patterns for target driven gates of the open segment under the single defect assumption. Later, the diagnosis flow first constructs the set of candidates based on a dictionary-based approach followed by an inject-and-evaluate analysis with the assistance of physical information to greatly reduce the candidate size. Experiments are conducted on ISCAS85 benchmark circuits and the results show that the proposed algorithm runs efficiently and can deduce nearly 1 candidate for each individual open-segment defect.