Army Vehicle Software Complexity Prediction Metric - Five Factors (Preprint)
Abstract
Army vehicle software provides mission critical complex functions. It interacts with complex electronics from multiple vendors and has unique software interfaces. The software structure complexity is influenced by many factors prior to software development. Understanding, predicting and resolving complexity of vehicle software prior to its development is a necessity for army mission success. Current complexity metrics are historical data distribution dependent. It focuses on software and its technical structure with no consideration of its influencing factors. Non-technical metrics related to software complexity are required to address diverse skill set including the management. Using non-technical metrics prior to software development enables management to spend resources early to resolve issues faster. In this paper, the authors propose five non-technical factor metrics based on the current software development process to predict future Army vehicle software complexity. Factor analysis and fuzzy logic techniques are used for developing, modeling and analyzing the software complexity prediction metric. The proposed metric is independent of software, programming language, and domain. This metric is data distribution independent.
Document Details
- Document Type
- Technical Report
- Publication Date
- Apr 14, 2010
- Accession Number
- ADA538879
Entities
People
- Harpreet Singh
- Macam S. Dattathreya
Organizations
- United States Army Tank Automotive Research, Development and Engineering Center