This paper introduces a new notion called Verification & Validation (V&V) Lifecycle Methodologies, examining what V&V is, and expanding its scope to the entire software lifecycle (much beyond traditional methods of software testing). V&V Lifecycle Methodologies are streamlined and efficient, yet highly holistic and concisely measured step-by-step threaded and sequenced approaches to predictively and deterministically building in software quality as it is developed across the entire software lifecycle. V&V is recast in a much more holistic definition and approach, based on a rapidly maturing V&V discipline, solid empirical data, and recent innovations in metrics-based software quality and reliability management and engineering.
What Is V&V?
According to the IEEE Standard Glossary of Software Engineering Terminology , V&V is defined as the process of determining whether:
- Requirements for a system or component are complete and correct.
- Products of each development phase fulfill the requirements or conditions imposed by the previous phase.
- Final systems or components comply with specified requirements.
What this means is that if deliverable software products, usually non-deliverable by-products of the software process, and final integrated systems all satisfy their specified requirements, successful V&V has occurred. A simpler way of saying the same thing, is that V&V is the process of successfully eliminating software defects from all software lifecycle artifacts on a stage-by-stage basis, before testing begins.
Myths & Misconceptions
Unfortunately, there are many different perceptions of what V&V is, and new definitions are being created across the world all of the time. The author has been retained to derive new definitions of V&V for international standardization despite the plethora of existing definitions. But, the challenge of successfully defining, managing, and using V&V doesn’t end there, as many myths and misconceptions abound about what V&V is (making it one of the most difficult and confusing software disciplines to be a part of):
- V&V, quality, and testing are often equated: V&V is the process of eliminating defects from software artifacts across the software lifecycle. Quality is a process of ensuring that all specified software processes have been properly executed. Testing is a process of dynamically executing software after it is complete. Many practitioners equate these three forms of processes, designate them as a single labor category, and often reduce V&V and quality to the mere act of testing.
- Quality and testing are often equated: Once again, quality processes and labor categories are often reduced to testing.
- Testing and V&V are often equated: V&V processes and labor categories are also often reduced to testing.
- Testing is believed to be sufficient: Even when V&V and quality processes and roles are well understood, testing itself is perceived to sufficiently replace them.
- V&V is often confused with IV&V: The role of V&V is an inherent part of software management and engineering, and even specialized labor categories internal to software lifecycles. IV&V is the process of V&V carried out by an independent group other than the principal software managers and engineers, for the purpose of ensuring objectivity. IV&V is often designated to be carried out by an independent third party contractor or wholly external organization, also for the purpose of ensuring objectivity. Unfortunately, IV&V is often, but not always, reduced to the process of testing.
- IV&V is believed to be better than V&V Lifecycle Methodologies: It is also popularly held that IV&V is a necessary part of mission critical software development. The notion being that IV&V can successfully be used to objectively eliminate defects, that software managers and engineers are unable to, based on an objective viewpoint.
At the very least, this paper will attempt to define what V&V and testing are, differentiating between them sharply. This paper will minimally alert the reader to the notion that quality, testing, V&V, and IV&V processes are not equal in definition, execution, or effectiveness. And, hopefully, this paper will begin to dispel some of these deeply ingrained myths and misconceptions concerning V&V.
Three unique V&V approaches will be identified and briefly explained here. The objective is to call out important software industry V&V perspectives. These three V&V approaches will be placed in the proper context, and the most appropriate and attractive form of V&V will be illuminated. Appropriateness and attractiveness will be defined from the perspective of the