MAMBA: A Measurement Architecture for Model-Based Analysis

Frey, Sören, van Hoorn, André, Jung, Reiner, Hasselbring, Wilhelm and Kiel, Benjamin (2011) MAMBA: A Measurement Architecture for Model-Based Analysis Technische Berichte des Instituts für Informatik, TR-1112 . Department of Computer Science, University of Kiel, Germany.


Download (385Kb) | Preview


Model-based measurement techniques are relevant in the field of software analysis. Several meta models for the specification of quantitative measures have been proposed. However, they often focus either on static or dynamic aspects of a software system. Nevertheless, considering reengineering activities often both dimensions reveal valuable complementary insights. Existing meta-models are also frequently bound to specific modeling languages, redefine underlying concepts for any new meta model, or provide only limited tool support for the automated computation of measurements from modeled measures.

We present MAMBA, an integrated measurement architecture for model-based analysis---both static and dynamic---of software systems, that can be specified by arbitrary Ecore-based modeling languages. MAMBA extends the Structured Metrics Meta-Model (SMM) by additional modeling features, such as arbitrary statistical aggregate functions and periodic aggregate functions, e.g., for dynamic analysis at runtime. To consider measurements for querying system models, we outline the MAMBA Query Language (MQL) that employs SMM measures. Furthermore, we provide tool support that applies the measures specified in an (extended) SMM model and can integrate raw measurements provided by arbitrary static and dynamic analysis tools to produce the desired measurement model.

We demonstrate the applicability of the approach based on three evaluation scenarios from different contexts: migration of software systems into the cloud, model-based engineering of railway control systems, and dynamic analysis for model-driven software modernization.

Document Type: Report (Research Report)
Research affiliation: Kiel University > Software Engineering
ISSN: 2192-6247
Date Deposited: 18 Feb 2012 06:05
Last Modified: 03 May 2013 12:35

Actions (login required)

View Item View Item

Document Downloads

More statistics for this item...