Applying Ant Colony Optimization in software testing to generate prioritized optimal path and test data

Software testing is one of the most important parts of software development lifecycle. Among various types of software testing approaches structural testing is widely used. Structural testing can be improved largely by traversing all possible code paths of the software. Genetic algorithm is the most used search technique to automate path testing and test case generation. Recently, different novel search based optimization techniques such as Ant Colony Optimization (ACO), Artificial Bee Colony (ABC), Artificial Immune System (AIS), Particle Swarm Optimization (PSO) have been applied to generate optimal path to complete software coverage. In this paper, ant colony optimization (ACO) based algorithm has been proposed which will generate set of optimal paths and prioritize the paths. Additionally, the approach generates test data sequence within the domain to use as inputs of the generated paths. Proposed approach guarantees full software coverage with minimum redundancy. This paper also demonstrates the proposed approach applying it in a program module.

I. Introduction

In software development life cycle, software testing covers almost 40-50% of the total development cost [1]. Testing cost can be reduced largely if the errors can be filtered earlier. However, test case generation, full path coverage, test result optimization are technically difficult to maintain manually. Random test data generation is used to select random inputs for Software Under Test (SUT) [2] [3]. But the tester is unable to find most of the errors because the testing requirements are not incorporated with the data generation process. So, automation of testing is desired both to enhance the quality of testing and reduce the testing cost [4].