A Standardized Software Reliability Measurement Methodology
Abstract
Current Air Force practice is to perform Operational Test and Evaluation (OT and E) for each new weapons system. In support of this, Headquarters Air Force Operational Test and Evaluation Center (HQ AFOTEC) is responsible for measuring both suitability and effectiveness. While suitability is adequately measured, the current effort only addresses hardware effectiveness, or at best, system effectiveness. Since tools and metrics are in place for software suitability assessments related to OT and E (for example, software maintainability), there should be some effective way of measuring the operational effectiveness of software. Currently, HQ AFOTEC/LG5 has a data collection tool for collecting software failure data to analyze software maturity. This thesis proposes that the LG5 software maturity database could be used as the baseline for a software reliability metric that would map to the finite time OT and E environment. This study does not predict software reliability, nor does it attempt to define what constitutes reliable software. Instead, this study evaluates software reliability measurement mapped to finite OT and E time frames (i.e.-failure per flight hour). This evaluation is conducted for several software reliability models, with two candidate models chosen based on the following criteria: predictive validity; capability; quality of assumptions; applicability to the finite-time environment; simplicity of design; diversity and applicability of output; and capability to use existing initial data.
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 01, 1991
- Accession Number
- ADA244184
Entities
People
- Joseph J. Stanko
Organizations
- Air Force Institute of Technology