Research Project:
Software Requirements Engineering and Design based on Message Sequence
Charts
Research Partners:
Stefan Leue (Principal
Investigator), Electrical & Computer Engineering, University of Waterloo,
<sleue@swen.uwaterloo.ca>.
Bran Selic, ObjecTime Limited,
<bran@ObjecTime.com>.
Research Team at Waterloo:
Mohammad Rezai (Post
Doc), <mrezai@swen.uwaterloo.ca>
Mohammad Zulkernine
Alumni:
Dr. Hanêne Ben-Abdallah
(Post Doc, 07/96 - 07/97), now with Faculté des Sciences Economiques
et Gestion, Université de Sfax, Tunisia
Remi Delon
Lars Mehrmann
Jennifer Hunt
Tuan Ngo
Funding:
Information Technology Reserach Centre (ITRC).
ObjecTime Limited.
National Science and Engineering Research
Council of Canada (NSERC), Collaborative Research & Development
Grant.
Duration:
Jan. 1996 - June 1998
Research Results
Overview (as of July 1997):
We are developing a tool to support the scenario-based design of
concurrent, object-oriented real-time systems based on Message Sequence
Charts. We developed algorithms that syntactically analyze ITU-T Z.120
hierarchical MSC specifications with respect to semantic ambiguities like
process divergence and non-local branching choices. Further work addresses
timing information in MSCs and reconciling the ITU-T Z.120 timer notation
with alternative proposed notations. We then suggest extending existing
timing consistency analysis algorithms (c.f. Dechter, Meiri and Pearl;
Alur, Holzmann and Peled) to branching and iterating Z.120-style MSC specifications.
The analysis algorithms are being implemented in the MESA toolset that
we are currently developing.
Reports and Papers (all of these are available as postscript file
through our compendium `Research
on Message Sequence Charts'):
-
MESA: Support for Scenario-Based Design of Concurrent Systems
-
Timing Constraints in Message Sequence Chart Specifications
-
Expressing and Analyzing Timing Constraints in Message Sequence Chart Specifications
-
Syntactic Detection of Process Divergence and non-Local Choice in Message
Sequence Charts
-
Syntactic Analysis of Message Sequence Chart Specifications
-
Architecture of a Requirements and Design Tool Based on Message Sequence
Charts
Tools:
-
MESA (MSC Editor, Simulator and Analyzer): Currently under design
and implementation, will eventually implement \
-
syntactic checks for process-divergence, timing consistency,
-
various semantics analysis functions, and
-
code synthesis (ROOM/UML-RT, SDL, Promela.
Implementation currently in Tcl/Tk and C++ under Solaris, but we intend
to eventually migrate to JAVA. For a preliminary description of the tool
see our technical report "MESA:
Support for Scenario-Based Design of Concurrent Systems". It is planned
to make a research version of this tool publicly available once a stable
version is available, probably towards the end of 1998. In case you would
like to receive a copy, please send us an email: <sleue@swen.uwaterloo.ca>.
-
xMSC (implemented by Tuan Ngo): A preliminary editor for basic MSCs
and HMSC graphs, checks an MSC specification for non-local choice (for
the algorithms see our TACAS'97 paper). Saves MSC specification in internal
format, and prints MSCs and HMSC to Postscript file. Implemented in Tcl/Tk
and C++ under SunOS. Predecessor of MESA, not publicly available.
Research Outline (Taken from original research proposal):
In this project we address
the question of what Message Sequence Charts (MSCs) inherently express,
and what users implicitly assume them to express;
the syntactic analysis of MSC specifications in order to detect ambiguities
and aspects of underspecification;
possibilities for extensions to notation to make implicit semantics assumptions
explicit;
the introduction of composition and refinement/abstraction concepts;
the synthesis of executable processes and related validation and verification
concepts (e.g., model checking of MSC specifications);
the usage of MSCs as `constraints'; and
the incorporation of real-time requirements and object-orientation concepts
into the language.
We aim at providing tool support for some or all of the ideas resulting
from the above research. Special attention is paid to the application of
the research results to the ObjecTime toolset, based on the ROOM methodology.
Go to the home page of the E&CE
department of the University of
Waterloo, or to the home page of the Software
Engineering Group (SWEN) of the E&CE department.
Last modified: July 17, 1998.
Stefan Leue <sleue@swen.uwaterloo.ca>