Optimal Release Planning and Software Reliability Modeling for Multi-Release Software

Abstract Reliability of Software has been major Issue in Software industry. Failure Free Code Development is to be achieved with in time of Software version upgrades. Although this give rise to numerous faults. Analyzing and resolving these Defects to achieve Software reliability is objective of Research. To overcome these defects Software domain has come up with software Reliability and Growth Models (SRGM’s). This Model predict faults that would occur and present count of defects in module. While Resolving this defects SRGM’s fail to achieve optimal efficiency in issues removal. Fault removal metric help product developer to evaluate performance and affectivity of software and overall work load computation. As such optimal planning and Reliability Modeling for Software system is active research area. Proposed System Analysis and model defects from previous Software Versions eliminating them. Modeled on this Concept a mathematical Analyzing and Modeling framework for multiple release of Software product is Designed and Developed. Research work achieves reliability with 2-dimension work flow applying testing for deployed product and Cobb- Douglas factor. Testing cannot be done continuously and research seeks to minimize testing, value minimization reliability maximization are outcomes of research. Optimal design is complicated non-linear equation to solve. Proposed system Deploys GA (genetic algorithm) to achieve Reliability as of survey outcomes presents GA to be best for Search and optimization, which is implemented interactively on real binary solution of population, with 4 genetic operations: crossover, mutation and reproduction with S shape Model at core. Functional dataset consisting test week, test time, resources and identified Faults. Optimal model solves and determines when to stop testing and predict with time newer version would be released, with removal of dead test cases and test coverage evaluation as added workings.