Input Space Coverage Matters

Area of Science:

• Software Engineering
• Computer Science
• Quality Assurance

Background:

• Software testing is crucial for assurance but often relies on subjective judgment.
• Structural coverage criteria enhance test completeness but may not detect faults from rare inputs.
• Existing test suites may lack comprehensive input diversity, leading to potential software vulnerabilities.

Purpose of the Study:

• To propose supplementing structural coverage measures with input space coverage.
• To highlight the importance of input space coverage for identifying software faults.
• To demonstrate the relationship between structural and input space coverage for verifying input models.

Main Methods:

• Reviewing existing structural coverage metrics in software testing.
• Analyzing the limitations of structural coverage in detecting input-related faults.
• Investigating the utility and application of input space coverage measures.
• Examining the correlation between structural coverage and input space coverage.

Main Results:

• Structural coverage, while valuable, is insufficient for complete software fault detection.
• Input space coverage measures can identify faults missed by structural coverage alone.
• A relationship exists between structural coverage and input space coverage metrics.
• Input space coverage provides a method for validating the adequacy of input models.

Conclusions:

• Structural coverage must be augmented with input space coverage for robust software assurance.
• Effective input space coverage enhances the identification of rare input-triggered faults.
• The integration of input space coverage validates the defined input model, improving overall software quality.