Arrays for combinatorial interaction testing: a review on constructive approaches

Interaction faults in component-based complex systems have been investigated from various aspects in the past few decades. The key issue is that, in a complex system, not only a single variable may cause the failure, but also the interactions among different variables. Combinatorial interaction testing has been established as one of the most fundamental and effective tools for identifying and locating the faulty interactions. This paper presents a review of the brief history, essential problems, and significant results on arrays for combinatorial interaction testing, including covering arrays, locating arrays, detecting arrays, and their variations. Specific approaches on explicit constructions for covering arrays involving information-theoretic methods are reviewed. Finally, we propose two improved bounds for studying the optimality of locating arrays.