Analysis and assessment of software reliability modeling with preemptive priority queueing policy

Abstract

Due to the growing complexity and size of software systems in modern society, the quality requirements for computer software have become ever more stringent. In the past, many software reliability growth models (SRGMs) have been proposed to evaluate and assess the reliability and quality of software systems. Some studies have demonstrated that both infinite server queueing (ISQ) and finite server queueing (FSQ) models can be applied to describe the fault detection process (FDP) and fault correction process (FCP) of software systems. However, it has also been noted that most ISQ and FSQ models assumed and obeyed the first come first served (FCFS) rule when removing the detected faults in FDP. In practice, the detected faults generally are classified into different levels of priority and those with higher priority should be fixed earlier. That is, high-priority faults have to be removed quickly to minimize their impact on software systems. Consequently, this assumption should be properly modified or adjusted. In this paper, we proposed a preemptive priority queueing (PPQ) model that considers both a finite number of debuggers and different priority levels. In our proposed PPQ model, faults assigned higher priority would be able to preemptively acquire resources already occupied by lower priority faults. Some numerical examples based on real failure data from different open-source and closed-source software are analyzed and discussed in detail. Experimental results show that the proposed PPQ model can provide more accurate estimation capability for software reliability, compared to traditional SRGMs.