Fault-Test Pair Ranking via Waveform Statistics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Mutation-based test program evaluation systems face challenges in efficiently identifying and ranking fault-test pairs that produce undetected waveform differences, which are crucial for detecting hidden problems in integrated circuit designs.
Innovation Solution
A method and system for ranking fault-test pairs based on waveform statistics, involving the calculation of relative differences between reference and faulty waveforms, and determining a ranking result from these differences to prioritize fault-test pairs for investigation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If mutation analysis is performed on a large number of mutations, then the quality of test suite evaluation is improved, but the time required for analysis increases significantly
Solution Approach 1:
The patent applies partial action by performing mutation analysis on a selected subset of mutations rather than all possible mutations. The system identifies and analyzes only those mutations that are most likely to reveal hidden problems, based on waveform difference criteria. This allows the test suite evaluation to maintain high quality while significantly reducing the total number of mutations analyzed, thereby decreasing analysis time.
2Loss of time
If the number of mutations is reduced to decrease analysis time, then the analysis time is reduced, but the ability to detect hidden problems decreases
Solution Approach 1:
The patent changes the parameter of mutation selection from random or uniform selection to selection based on waveform difference characteristics. By analyzing waveform statistics and identifying mutations that produce significant waveform differences, the system prioritizes mutations that are more likely to detect hidden problems. This parameter change ensures that even with a reduced number of mutations, the detection capability is maintained or improved.
Solution Approach 2:
The patent applies local quality by focusing analysis resources on specific mutations that exhibit particular waveform characteristics indicative of hidden problems. Rather than treating all mutations equally, the system identifies mutations with localized waveform differences that are more likely to reveal test suite deficiencies, and concentrates analysis efforts on these high-value mutations.
3Reliability
If all fault-test pairs are analyzed in detail, then the detection of hidden problems is improved, but the complexity of the analysis process increases
Solution Approach 1:
The patent segments the analysis process into distinct phases: initial waveform comparison, identification of mutations with significant waveform differences, and detailed analysis of only those segmented cases. This segmentation allows the system to handle large numbers of mutations efficiently by breaking down the complex analysis into manageable stages, reducing overall process complexity while maintaining detection reliability.
Data Source
AI summary
Ranking of fault-test pairs is performed using first and second multitudes of waveform statistics. The first multitude of waveform statistics includes first value-change information regarding variations in logics HIGH and LOW for each bit of each reference output resulting from a test run of the design code. The second multitude of waveform statistics includes second value-change information regarding variations in logics HIGH and LOW for each bit of each faulty output resulting from a test run of the design code injected with a fault. Relative differences between the first and second multitudes of waveform statistics for each bit of each faulty output with respect to the corresponding reference output are determined. A waveform difference based on the relative differences for each signal of each faulty output is determined. A ranking result of fault-test pairs is determined according to the waveform differences of the faulty outputs.