AbstractsCommunication

Evaluation of Model Based Testing

by Yasir Malik




Institution: Blekinge Institute of Technology
Department:
Year: 2010
Keywords: telekommunikation; telecommunications; model based testing; software testing; qtronic conformiq.
Record ID: 1333800
Full text PDF: http://www.bth.se/fou/cuppsats.nsf/6753b78eb2944e0ac1256608004f0535/99e7aeac74c2614dc1257727004b94e6?OpenDocument


Abstract

In today’s telecommunication market, the increased complexity in the system and short release cycle of a product is becoming a challenge for the product testers. Due to increased competition in the telecom sector, a high reliability of complex software is always demanded by the customer, along with the cost reduction, from his contractor. A product has to be tested again if a small modification or some extra functionality is added in it. As testing is traditionally performed manually, which cannot assure that the software is tested using all possible combination of inputs. Therefore, to enhance the reliability of tests, there is a need of techniques which can improve the manual way of testing and assure the high performance and evaluation of the product. Clearly, the test automation techniques are getting more consideration due to its benefits. Therefore, in this thesis work an evaluation is made on model based testing (MBT) using Qtronic by Conformiq. Qtronic is a tool for automatic test case design that is driven by ‘design models’. A simplified automatic teller machine (ATM) client-server system is used initially as a system under test (SUT), which is implemented in Java. Qtronic modeling language (QML) is used to design a model of ATM using finite state machines (FSM) notation. The ‘design model’ is a description of the intended behavior of the system on some level of abstraction. Qtronic designs test cases for a system automatically when it is given a ‘design model’ of a system as an input. The complexity of the test object is increased incrementally to evaluate how well suited Qtronic is for incremental design and how changes in a test object affects model based testing in broad implementation. Furthermore, an experiment is also performed to evaluate the test generation time of Qtronic, by moving the ‘core logic’ of the model to the test harness. However, it is recommended that a larger and more complex test object should be used to evaluate the model based testing using Qtronic.