Projects   
   
This page gives an overview of the projects I have worked on, or am still working on.
Explicitly modelled client, server, and network for the MvK
This work removes the need for coded libraries to facilitate HTTP communication between the MvK clients and server. My contribution consists of:
  • A SCCD socket library, which allows a mapping between SCCD events and socket operations
  • Model a small HTTP server in SCCD, including server-initiated communication, multi-client, ...
  • Model a small HTTP client in SCCD, including server-initiated communication, event-based interface, ...
  • Model the communication protocol between client and server (including network) in DEVS
  • Generate code from the HTTP client and server SCCD models
  • Performance evaluation using the DEVS models
  • Modify the MvK frontend (by Simon Van Mierlo) to use the explicitly modelled network
It consists of a small report.
DEVS debugger
Together with Simon Van Mierlo, I developped a DEVS debugger in AToMPM. My contribution to this work consists of:
  • Lift the modal part out of PythonPDEVS
  • Model the modal part in SCCD (Statecharts + Class Diagrams)
  • Implement better realtime simulation
  • Include debugging operations
It consists of a technical report and the DEVS debugger.
Python(P)DEVS
For my first research internship, I increased the performance of PythonDEVS. My contribution consists of:
  • Literature study about DEVS
  • Prototype some optimizations in PythonDEVS
  • Benchmark the effect of the optimizations
A detailed overview can be found in my report.
In my second research internship, PythonDEVS was reworked to a prototype of distributed PythonPDEVS. Additions of this are:
  • Literature study about distributed simulation
  • Prototype some further optimizations in PythonPDEVS
  • Benchmark the effect of the optimizations
  • Implement distributed simulation using Time Warp
A detailed overview of my work can be found in my report.
Finally, in my Master's thesis, I finalized my work on PythonPDEVS with the following changes:
  • Stabilize simulation algorithms
  • Activity information for allocation and migration
  • Domain-specific hints can be supplied by the model
  • Add further simulation options, such as realtime simulation, modularised tracing, and modularised scheduling
A global overview of PythonPDEVS can be found in my Master's thesis. All these and previous changes are present in PythonPDEVS.
Modelling the DEVS formalism, with symbolic flattening
For the course Model-Driven Engineering, I expanded the AToM3 DEVS formalism (by Bill Song) and the Modelica symbolic flattening (by Bin Chen). My contribution consists of:
  • Update the AToM3 formalism with the newer version of PythonDEVS
  • Merge XML tracing in PythonDEVS
  • Reimplement the symbolic flattening
  • Performance evaluation of symbolic flattening for DEVS models
It consists of a report and the material
Logisim to DEVS translator
In this work, I continued the work of the Logisim to Verilog translator by Naomi Christis. My contribution consists of:
  • Update logisim to verilog translator to also support logisim 2.7.1
  • Make a library of primitive models in DEVS
  • Modify back-end of translator to generate DEVS
  • Modify back-end of translator to generate trace files (VCD and XMLTracePlotter formats)
It consists of my Bachelor's thesis and the translator.
Maintained by Yentl Van Tendeloo. Last Modified: 2015/09/11 09:22:28.