The Software Life Cycle (SLC) often comprises a complex sequence of processes, each with many subparts where various execution decisions throughout the pipeline can greatly affect the success or failure of a given project. Some of the most important decisions involve the allocation of scarce resources throughout the SLC, which are often based on estimations about future market demand and various extraneous factors of high stochasticity. Despite numerous efforts in standardization, many projects are still highly dependent on the subjective aptitude of individual managers, who may in turn rely on ad hoc techniques rather than standardized and repeatable ones. The results can be unpredictability and undue reliance on specific individuals.
This paper considers imposing a mathematical framework on two of the key aspects of SLC: Deciding how to dynamically allocate available resources throughout the development pipeline, and when to stop further work on a given task in light of the associated Return On Investment (ROI) metrics. In so doing, the software development process is modeled as a problem in New Product Development (NPD) Management, which can be approached using control theory and stochastic combinatorial optimization techniques. The paper begins by summarizing some of the previous developments in these fields, and proposes some future research directions for solving complex resource allocation problems under stochastic settings. The outcome is a formal framework that when combined with competent Configuration Management techniques, can rapidly achieve near-optimal solutions at each stage of the SLC in a standardized manner. / text
| Identifer | oai:union.ndltd.org:UTEXAS/oai:repositories.lib.utexas.edu:2152/ETD-UT-2011-12-4623 |
| Date | 27 February 2012 |
| Creators | Mozano, Ashton |
| Source Sets | University of Texas |
| Language | English |
| Detected Language | English |
| Type | thesis |
| Format | application/pdf |
Page generated in 0.0018 seconds