Waterfall Model Simulation: A Systematic Mapping Study

TL;DR

The paper addresses the gap in understanding how the traditional waterfall software lifecycle has been represented through computational simulation. Using a rigorous systematic mapping across six major databases (2000–2024) and a PRISMA‑guided process, it identifies five studies that explicitly simulate waterfall, finds discrete‑event simulation as the dominant approach, and reveals limited tool disclosure and fidelity to Royce’s original phases. It shows a global but sparse research footprint, with some studies comparing waterfall to agile or hybrid approaches, and none reproducing the complete seven‑phase Royce model. The authors propose clearer reporting of model structure and tool usage, and encourage open‑source tooling and reproducible artifacts to strengthen future work. Overall, the study provides a consolidated view of waterfall simulation research and outlines concrete directions for improving transparency and methodological rigor in this niche area.

Abstract

This paper systematically maps peer-reviewed research and graduate theses/dissertations that explicitly simulate the waterfall model. Following Petersen's mapping guidelines and Kitchenham's systematic literature review practices, major databases (ACM Digital Library, IEEE Xplore, Scopus, Springer, Google Scholar, and Web of Science) were searched for studies published between 2000-2024 using the title query ("simulation" OR "simulating") AND "waterfall". A PRISMA workflow guided the screening process, and approximately 9% of retrieved records met the inclusion criteria. A repeated extraction process captured methods, tools, venues, geography, publication years, comparative scope, and fidelity to Royce's original model; findings were synthesized thematically. Discrete-event simulation dominates (80%) compared to system dynamics (20%). Reported tools center on Simphony.NET (40%) and SimPy (20%), while 40% of studies omit tool details, limiting reproducibility. Research is distributed across Italy, Lebanon, India, Japan, and the United States; publication venues include 60% journals and 40% conferences. Sixty percent of studies are comparative, while 40% model only the waterfall approach. No study reproduces Royce's original model; all employ adaptations. The paper concludes by presenting a consolidated view of waterfall simulation research and recommending clearer model reporting, fuller tool disclosure, and wider adoption of open-source platforms.

Figures (5)

• Figure 1: Three formulations of the waterfall model (adapted from Royce royce1987): (a) single pass with strictly sequential phases and minimal feedback; (b) two-pass formulation in which the full lifecycle is executed twice; and (c) single pass with explicit feedback loops to the immediately preceding phase(s).
• Figure 2: PRISMA 2020 Flow Diagram Illustrating the Study Selection Process.
• Figure 3: Pie Charts Illustrating (a) the Type of Approach Used for the Simulation and (b) the Tool That was Used.
• Figure 4: Pie Charts Illustrating (a) the Country of Origin of the Authors of the Paper and (b) Whether the Publications Appeared as Conference Proceedings or Journal Articles.
• Figure 5: The Cumulative Number of Publications by Year.