Comparative Analysis of IPOG Strategy Variants for Enhanced Combinatorial T-way Testing Efficiency

Abstract

Enhancing software quality and precision requires thorough testing of various configuration aspects.  Combinatorial interaction testing has emerged as a potent method, utilizing strategies like Generalized Input Parameter Order (IPOG), which employs a one-parameter-at-a-time (OPAT) approach.  However, the challenge of combinatorial explosion, where the number of inputs grows exponentially with test cases, remains significant.  To address this, numerous strategies have been proposed, yet consistently generating optimal tests covering each t-way interaction efficiently remains elusive.  This paper introduces a novel variant of the IPOG strategy, termed the enhancing IPOG strategy for uniform interaction testing (eIPOG).  We investigate the application of the harmony search algorithm in introducing optimal combinatorial interaction t-way test suites through eIPOG.  Evaluating the behavior of such strategies involves assessing their effectiveness in minimizing the test suite size while maximizing coverage.  Through experiments on established benchmarking configurations and statistical analysis, we compare eIPOG with other IPOG-based strategies.  Both approaches show promise, highlighting the efficacy of the OPAT approach, with eIPOG demonstrating particularly competitive results in nearly every case.