Using simulation to evaluate global software development task allocation strategies

Abstract

We describe a hybrid computer simulation model of the software development process that is specifically architected to study alternative ways to configure global software development (GSD) projects, including phase‐based, module‐based, and follow‐the‐sun allocation strategies. The model is a hybrid system dynamics and discrete‐event model. In this paper, test cases have been developed for each allocation strategy, and project duration is computed for each configuration under a range of plausible assumptions for key parameters. The primary finding is that although under ideal assumptions follow‐the‐sun is able to produce impressive reductions in time‐to‐market, under more realistic assumptions the reverse is true, thus corroborating findings by other researchers. We also conducted a factorial design to examine the impact of GSD factors including distance, culture, language, trust, and time zone on project duration under different task allocation strategies. The analysis reveals that different factors affected the performance of the selected allocation strategies in unique ways. These findings show how the unique ability of our GSD model to represent detailed development processes and work artifact transfer allows researchers to address challenging questions that are critical to GSD project success. Copyright © 2007 John Wiley & Sons, Ltd.

A Reinforced Hungarian Algorithm for Task Allocation in Global Software Development

“…The successful allocation rate of the expansion matrix method decreases when the number of tasks increases It shows that RHA can effectively allocate tasks to development sites with the lowest cost when the number of tasks changes. In this section, we briefly review the existing task assignment methods [13][14][15][16][17][18][19][20][21][22][23][24][25][26] in global software distribution.…”

“…Amrit proposes a methodology to test a hypothesis based on how social networks can be used to improve coordination in software industry [21]. Setamanit describes GSD-a hybrid computer simulation model of the software development process in order to identify the practices of work distribution and try to classify criteria [22]. Marques presents a domain ontology to represent concepts related to task allocation in distributed teams in his paper.…”

Finding robust solutions in requirements models

“…There exist many powerful requirements analysis tools including continuous simulation (also called system dynamics) [1,65], state-based simulation (including petri net and data flow approaches) [3,29,47], hybrid-simulation (combining discrete event simulation and systems dynamics) [19,46,63], logic-based and qualitative-based methods [7, chapter 20] [37], and rule-based simulations [52]. One can find these models being used in the requirements phase (i.e.…”

Developing an Inter-site Coordination Index for Global Software Development

“…Using simulation studies on a hybrid system dynamics and discrete-event model, Setamanit et al [18] analyze the impact of a number of factors on project duration under four possible task allocation strategies namely single-site, module-based, phasebased, and follow-the-sun. Espinosa and Carmel [6] use a mathematical model to illustrate the impact of time-zone differences on project completion time.…”