Efficient software testing protocols

  • Authors:
  • Brett Stevens;Eric Mendelsohn

  • Affiliations:
  • Department of Mathematics, University of Toronto, 100 St. George St. Toronto ON M5S 3G3;Department of Mathematics, University of Toronto, 100 St, George St. Toronto ON M5S 3G3

  • Venue:
  • CASCON '98 Proceedings of the 1998 conference of the Centre for Advanced Studies on Collaborative research
  • Year:
  • 1998

Quantified Score

Hi-index 0.00

Visualization

Abstract

Testing software against failure rapidly, efficiently and comprehensively is of vital importance to software programmers and designers. Complete testing is far too slow and costly. Randomized testing can suffer from deficiencies of the pseudo-random number generator used. Testing all pairwise input interactions is a relatively new idea that resolves this problem. It is extraordinarily efficient; the number of test cases only grows as the log of the number of inputs. Recent evidence shows that the vast majority of all software failures are single or pairwise input interactions. A test protocol that covers all pairwise interactions of inputs is called a covering array. For applications, the central issue is to be able to produce small covering arrays with given parameters on demand. This paper presents a number of methods to generate efficient pairwise testing protocols. A software package using these methods to generate test protocols is also discussed.