Run-time software configuration for mobile devices using an evolutionary quantifiable deployment model

  • Authors:
  • Jakob Sverre Lovstad;Peter Henry Hughes

  • Affiliations:
  • IDI, NTNU, Trondheim, Norway;IDI, NTNU, Trondheim, Norway

  • Venue:
  • WOSP '08 Proceedings of the 7th international workshop on Software and performance
  • Year:
  • 2008

Quantified Score

Hi-index 0.00

Visualization

Abstract

Current proposals for performance evaluation during development of computer systems rely on the capture of performance information from various software development models such as UML diagrams. The requirements for performance are taken into consideration at some point in the development process and evaluated using a performance model. We propose an enhanced approach using a separate, quantifiable deployment model that evolves in step with system development. This helps keep performance requirements in focus from beginning to end of the life-cycle. The system implemented to demonstrate this approach is a computer game for mobile devices that is configured dynamically for different target devices. The development process must provide for deferred, automatic performance testing and configuration of the application on a variety of devices, after the design and implementation stages are complete. The evolutionary deployment model, culminating in a run-time performance evaluation, is based on the Structure and Performance modelling paradigm SP.