EWD Metamodeling Environment

Duration: Current

Project Goals

To model complex embedded systems, one needs a modeling framework that allows expressibility and fidelity to capture multiple models of computation (MoC)s and their interaction. However, none of the available Multi-MoC frameworks such as Ptolemy and SystemC extensions are as easily customizable by users, as metamodeling frameworks proposed in software engineering, such as GME. Also, none of these are multi-targeted in the sense that several special purpose models for synthesis, analysis or verification can automatically be generated from the framework. We therefore, present EWD environment and methodology, based on GME (Generic Modeling Environment), which allows users to customize the MoC specific visual modeling syntax, allowing design-time automatic syntactic and semantic checks on the models for conformance to their metamodel. Furthermore, we have built parsing and translation tools to save such models in an XML based interoperability language (IML). The IML format is in turn translated into Standard ML (SML) or Haskell models, which can then be executed and analyzed either by our existing model analysis tools, or the ForSyDe environment. We also generate SMV code from the XML format to obtain verification models. In summary, EWD is a novel visual multi-MoC modeling environment with interoperable intermediate representation of models, and has facilities to enforce syntactic and semantic checks and the ability to generate code for multiple target environments for further analysis and execution. The flexibility and interoperability comes from the meta-modeling environment GME, and use of XML based intermediate representation.

Design Flow

EWD Downloads

1. Multi-MoC Metamodel (Multi-MoC) [Install GME and then download and register this metamodel]

2. IML Syntax

3. xmlTree Data structure and Translation Streams

Snapshots

· An abstract View of the Multi-MoC

· Class Diagram of the xmlTree Data structure

Examples

· Modeling of an Adaptive Amplifier in EWD

· Modeling a Digital Equalizer in EWD

· Modeling of a Power State Machine in MoC++

Technical Reports

1. Deepak Mathaikutty, Hiren Patel, Sandeep Shukla and Axel Jantsch, EWD: A Metamodeling Driven Customizable Multi-MoC System Modeling Environment FERMAT Technical Report 2004-20, 2004.

Journal

1. Deepak A. Mathaikutty, H. Patel, S. Shukla and A. Jantsch, EWD: A Metamodeling Driven Customizable Multi-MoC System Modeling Environment, Submitted to IEEE Transactions on Computer-Aided Design, 3rd review.

Publication

1. Deepak Abraham Mathaikutty, Functional Programming and Metamodeling frameworks for System Design, Master’s Thesis, May 2005.

SML-Sys Framework

ForSyDe

Talks

1. Bridging the UML to SystemC gap, SystemC Panel, FDL 2004.

Personnel at Virginia Tech

PI:                                  Sandeep Shukla
COLLABORATOR:     Axel Jantsch
GRA:                             Deepak Mathaikutty
GRA:                             Hiren Patel