Citation: Kocatas, A.T.; Dogru, A.H.
Enhancing UML Connectors with
Behavioral ALF Specifications for
Exogenous Coordination of Software
Components. Appl. Sci. 2023, 13, 643.
https://doi.org/10.3390/
app13010643
Academic Editor: Sanjay Misra,
Robertas Damaševiˇcius and Bharti
Suri
Received: 23 November 2022
Revised: 24 December 2022
Accepted: 29 December 2022
Published: 3 January 2023
Copyright: © 2023 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
Article
Enhancing UML Connectors with Behavioral ALF Specifications
for Exogenous Coordination of Software Components
Alper Tolga Kocatas
1,2,
* and Ali Hikmet Dogru
2,3
1
Aselsan Inc., 06750 Ankara, Turkey
2
Department of Computer Engineering, Middle East Technical University, 06800 Ankara, Turkey
3
Department of Computer Science, University of Texas at San Antonio, San Antonio, TX 78249-0667, USA
* Correspondence: tolga.kocatas@metu.edu.tr or kocatas@aselsan.com.tr
Abstract:
Connectors are powerful architectural elements that allow the specification of interactions
between software components. Since the connectors do not include behavior in UML, the components
include the behavior for coordinating the components, complicating the designs of components and
decreasing their reusability. In this study, we propose the enrichment of UML connectors with
behavioral specifications. The goal is to provide separation of concerns for the components so
that they are freed from coordination duties. The reusability of the components will increase as
a result of such exogenous coordination. Additionally, using the associated behaviors, we aim
to resolve the ambiguities that arise when n-ary connectors are used. We use a series of QVTo
transformations to transform UML models that include connector behaviors in ALF specifications
into UML models which include fUML activities as connector behavior specifications. We present a set
of example connectors specified using the proposed method. We execute the QVTo transformations
on the example connectors to produce models that represent platform-independent definitions of the
coordination behaviors. We also present and discuss cases from real-life large-scale avionics software
projects in which using the proposed approach results in simpler and more flexible designs and
increases component reusability.
Keywords: ALF; behavior; connector; fUML; model transformation; port; QVT; QVTo; UML
1. Introduction
Using models in software development to cope with complexity and increase quality
is a proven approach employed by model-driven software development (MDSD). In MDSD,
models do not only constitute documentation but they are also considered equal to code,
as their implementation is automated [
1
]. MDSD saves time by increasing productivity
and reducing implementation errors. The approach has become more popular since the
introduction of the Unified Modeling Language (UML) [
2
]. Due to its ability to precisely
define the structural and behavioral aspects of the software systems, UML is widely used in
safety-critical real-time embedded software development. UML specification has matured
over the years. Extensions for real-time software development and systems engineering are
introduced. Open-source and commercial tool support for UML also increased significantly.
Although there are languages for software architecture description, the current ver-
sion of UML covers the core concepts, such as components, ports, and connectors. UML
provides various diagrams for modeling the structural and behavioral aspects of soft-
ware systems. Component diagrams describe the structures of the components and the
communication among them using ports and interfaces. The ports define interfaces of
software components and their interactions with the environment using required and
provided interfaces. Starting with UML 2.0, ports can also be included in the designs of
encapsulated classifiers.
Among various advantages of using ports, the most important one is to decouple
components and encapsulated classifiers from their environments. Second, ports provide
Appl. Sci. 2023, 13, 643. https://doi.org/10.3390/app13010643 https://www.mdpi.com/journal/applsci
