Ranking of software reliability growth models using Bacterial Foraging Optimization Algorithm

Developing new software is an easy task but is really very difficult when it comes to change existing software, because the new changes can affect the old functionalities of the software, hence making the software unreliable. It doesn´t matter how much testing is done, but the correct result is known once the product is delivered to the customer. Hence, we need to know whether our products are reliable or not before delivering them to customers. This can be achieved by Software Reliability Models. In this paper, Bacterial Foraging Optimization Algorithm (BFOA) technique is operated on Tandem Computer data set to estimate parameters of thirteen software reliability growth models based on Least Square Estimation fitting technique. These models were then compared against eleven comparison criteria like BIAS, Variance etc. Then Distance Based Approach (DBA) is applied to rank all these models based on the comparison criteria. Among the thirteen models under study inflection s-shaped model and goel-okumoto model were ranked one and two respectively.