Toward a Discipline of Scenario-Based Architectural Engineering

Abstract

Software architecture analysis is a cost-effective means of controlling risk and maintaining system quality throughout the processes of software design, development and maintenance. This paper presents a sequence of steps that maps architectural quality goals into scenarios that measure the goals, mechanisms that realize the scenarios and analytic models that measure the results. This mapping ensures that design decisions and their rationale are documented in such a fashion that they can be systematically explored, varied, and potentially traded off against each other. As systems evolve, the analytic models can be used to assess the impact of architectural changes, relative to the system's changing quality goals. Although scenarios have been extensively used in software design to understand the ways in which a system meets its operational requirements, there has been little systematic use of scenarios to support analysis, particularly analysis of a software architecture's quality attributes: modifiability, portability, extensibility, security, availability, and so forth. In this paper we present a unified approach to using scenarios to support both the design, analysis and maintenance of software architectures, and examples from large-scale software development projects where we have applied the approach. We also present a tool, called Brie, that aids in: scenario capture, mapping scenarios to software architectures, and the association of analytic models with particular portions of architectures. The approach that we have devised, and that Brie supports, is a foundation for a discipline of architectural engineering. Architectural engineering is an iterative method of design, analysis and maintenance where design decisions are motivated by scenarios, and are supported by documented analyses.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jan 01, 2000
Accession Number
ADA469331

Entities

People

  • Rick Kazman
  • S. J. Carriere
  • Steven G. Woods

Organizations

  • Carnegie Mellon University

Tags

Communities of Interest

  • Energy and Power Technologies
  • Materials and Manufacturing Processes
  • Space

DTIC Thesaurus Topics

  • Abstracts
  • Air Traffic
  • Air Traffic Control Systems
  • Complex Systems
  • Computer Programming
  • Computer Programs
  • Computing System Architectures
  • Control Systems
  • Databases
  • Engineering
  • Failure Mode And Effect Analysis
  • Infrastructure
  • Markov Models
  • Models
  • Reliability
  • Software Design
  • Software Development

Fields of Study

  • Computer science
  • Engineering

Readers

  • Software Engineering.
  • Team-Based Human-Centered Cognitive Task Decision Making and Information Performance.