China
Africa
Explore chapters and articles related to this topic
The Technology in the Standard
Published in Klaus Diepold, Sebastian Moeritz, Understanding MPEG-4, 2012
Klaus Diepold, Sebastian Moeritz
An entirely different type of information is compiled when testing for conformance. Conformance testing is done by a company to check if its product is compliant with a target profile and level combination relevant for its business. That is, the engineers in the company will have built chips or programmed software according to the book. After they are done, the result will need to be tested for conformance. To this end, MPEG offers Part 4 of the standard, which describes in detail how the tests need to be done, what figures need to be measured, and what the results should look like in order to successfully verify the compliance with the standard. Furthermore, part 4 of the MPEG standard also contains bit streams that have been verified in the course of the standardization work to be legal MPEG bit streams. The engineers in the company can now take those bit streams and decode them with the newly implemented decoder tool. If the deviation of the decoded material from the specification is below a certain threshold, then the decoder is considered to be compliant. The result of conformance testing is a black and white answer to the question of whether a product is compliant with a conformance point in the standard. Hence, conformance testing is a self-certification process for a company, as there is no legal entity to officially distribute certificates of conformance.
Design
Published in Miroslav Popovic, Communication Protocol Engineering, 2018
The main goal of TTCN is to enable conformance testing of software products. Conformance testing is used to check if the software product is compliant with its specification. Conformance test cases deal only with the external behavior of a software product. Actually, conformance testing is based on the application of the “black box” principle. The product must interact with its environment as specified. Its internal structure and behavior are not significant.
Parallel algorithms for reducing derivation time of distinguishing experiments for nondeterministic finite state machines
Published in International Journal of Parallel, Emergent and Distributed Systems, 2018
Khaled El-Fakih, Gerassimos Barlas, Mustafa Ali, Nina Yevtushenko
In FSM-based (conformance) testing, we have the FSM specification and a black-box FSM Implementation Under Test (IUT) about which we lack some information. This information can be deduced by conducting experiments on the IUT. An experiment consists of applying input sequences to the IUT, observing corresponding output responses and drawing a conclusion whether the machine under test conforms to its specification. An experiment (or a test) is adaptive if the selection of the next input to be applied to an IUT is based on the observed outputs of the IUT to the previous inputs, and an experiment is preset if the input sequences are fixed a priori [6,7]. In FSM-based conformance testing, mutation testing [8–10], and fault diagnosis [11–13], special input sequences, called distinguishing sequences, are usually used for different reasons. For instance, in conformance testing, distinguishing sequences are applied for checking the conformance of two machines, or of two states of the same machine. In mutation testing, given a specification FSM, one can generate (FSM) mutants of the given specification such that each of these mutants represents a potential faulty IUT, and then derive tests to distinguish these mutants from the given specification. An FSM mutant is usually derived from the given specification by altering the output(s) and/or destination states of some transitions. In FSM-based fault diagnosis, distinguishing tests are used to locate the faulty FSM IUT; that is, these tests can be used to distinguish the derived mutants from each other in order to locate the mutant that has the same behavior as the given faulty IUT.