Principles of Built-In-Test for Run-Time-Testability in Component-Based Software Systems

  • Authors:
  • Jonathan Vincent;Graham King;Peter Lay;John Kinghorn

  • Affiliations:
  • Southampton Institute, East Park Terrace, Southampton, Hampshire, SO14 OYN, U.K. jonathan.vincent@solent.ac.uk;Southampton Institute, East Park Terrace, Southampton, Hampshire, SO14 OYN, U.K. graham.king@solent.ac.uk;Philips Semiconductors, Millbrook Ind. Est., Southampton, Hampshire, SO15 ODJ, U.K. peter.lay@philips.com;Philips Semiconductors, Millbrook Ind. Est., Southampton, Hampshire, SO15 ODJ, U.K. john.kinghorn@philips.com

  • Venue:
  • Software Quality Control
  • Year:
  • 2002

Quantified Score

Hi-index 0.00

Visualization

Abstract

This paper examines the motivations and expectations of Built-In-Test (BIT) techniques for Run-Time-Testability (RTT) in component-based software systems. The difficulties associated with testing and integrating fully encapsulated components lead to a requirement for testing interfaces. The format of these interfaces is explored at a high level of abstraction, and some possibilities for Built-In-Test (BIT) are described. BIT is concerned with the detection of error conditions arising internally to a component, or arising from erroneous component interactions, and the propagation of these error conditions to a system component having responsibility for error handling and/or recovery. The implications for testability, reliability and maintainability are discussed, and it is concluded that BIT offers potential for improved product quality. Whilst the proposed approach is considered appropriate for a wide range of software systems, issues related to real-time systems, such as deadlock and timing constraints are of particular interest.