Localization extension of COSMIC-FFP method on complex algorithms

Functional size measurement is an important feature property of software metrics. As the core of software, algorithms are difficult to estimate due to different technologic methods and processing flows. This paper establishes the localization extension measurement principles on COSMIC-FFP FPA for complex algorithms and proposes function recognition principles, data group recognition principles and measurement steps. The paper also analyses the influence of complex data on software metrics. Measurement results of algorithms written by different languages in signal processing, characteristic pick-up and location tracking has indicated that the localization extension method excels standard method.