Students Projects (2000-2007)[All project ...]
Thanks to the students who have accepted to work with us and do their best to produce, the following MSc and research projects:
Project 1: Collaboration in the Virtual Laboratory for e-Sciences (MSc project, October 2004)
Project 2: Performance Models for Virtual Laboratory Modules (MSc project, May 2005)
Project 3: Module Deployment & Management within the Grid-based Virtual Laboratory for e-Science (May 2005)
Project 4: A Comparison of Interpolation Techniques for Spatial Data Prediction (MSc project, June 2006)
Project 5: Web Services Discovery on Gnutella (decentralized P2P networks) (MSc project, September 2006)
Project 6: Simulation for Scheduling Parameter Sweep Tasks on Global Grids (MSc project, September 2006)
Project 7: Performance Modeling of UML Designs: An approach based in Petri Nets(MSc project, August 2007)
Project 8: Problem solving environment for Medical image analysis (MSc project, August 2007)
Project 9: Streaming data between Web Services. Comparison of streaming protocols over a stream-enabled Web Service (Profile Project, September 2007)
Project 1: Collaboration in the Virtual Laboratory for
The Virtual Laboratory for e-Sciences seeks to provide users with a collaborative environment in which they will be able to work together across time and space while using Grid technology. In
this MSc project we will define the requirements for collaboration in the VL-e. An in depth study of the Userlist, Instant Messenger and Telepointer has been done and a Grid Service based architecture has been designed for the first two.
Alexander de Ridder
Project 2: Performance Models for Virtual Laboratory
This MSc thesis describes techniques and proposes an architecture for semi-automatically computing dynamic performance models for modules, the components to build applications in the Virtual Laboratory Amsterdam. Our approach succeeds at modeling the most important factors (execution time and memory usage) that affect performance for the modules and enables to explore the interactions between a target architecture and application characteristics.
May 20, 2005
Project 3: Module Deployment & Management within
the Grid-based Virtual Laboratory for e-Science
The objective of this MSc thesis, is to propose an automated module deployment and module management system for VL-e. An architecture for module deployment is presented. Furthermore, the aspect of module management is researched from a scientific perspective. An important task of module management is to remove the least needed module in case there is no more space left on a VL-e resource. Trace-driven simulations are performed on real and artificial data to find the best VL module replacement strategy. The strategies used here, were originally designed for implementation in hardware and web caching systems.
Project 4: A Comparison of Interpolation Techniques for
Spatial Data Prediction
The objective of this MSc thesis is to provide a comparison between eight interpolation and regression techniques through an automated model that is able to produce different species distribution and abundance maps based on information obtained from sample sites. This automated model will be run under the Grid-based Virtual Laboratory Amsterdam (VLAM-G) which provides access to geographically distributed resources. This will provide extra accessibility and will speedup the performance of the automated model.
Project 5: Web Services Discovery
on Gnutella (decentralized P2P networks)
The Goal of this MSc thesis we propose a simulation framework which can be used to achieve this goal based on GnutellaSim, we also extend one of the most commonly used semantic matching algorithms using elements from the OWL-s description language, allowing a wider range of matching possibilities.
Project 6: Simulation for Scheduling Parameter
Sweep Tasks on Global Grids
Parallel and distributed systems are now days moving towards to the concept of Grid computing. The globalization of dynamic and heterogeneous resources connected via Internet and shared by different users is now days a reality. The new power that rises gave the ability to scientists to turn into type of studies that was before for them a utopia. Parameter sweep studies are such a type of study. This MSc thesis will try to investigate the impact that will have the consideration of communication cost within a parameter sweep task workflow. For that reason a scheduling algorithm will be analyzed and modified.
Project 7: Performance Modeling of UML Designs:
An approach based in Petri Nets
This MSc thesis presents an approach for performance modeling based on UML designs, using the use-case, deployment and sequence diagrams. The approach is based on a transformation from UML to Reference Nets (a high-level Petri Net formalism) and uses simulation as means of analysis. Over this basis we propose a methodology for Performance Engineering that starts at the design stage and progresses throughout the service lifecycle. So that common solutions and best-practices can be encapsulated for use by non-specialists. From the notion of a Software Pattern that can be executed for performance analysis we them attempt to define the concept of Executable Software Pattern.
Pedro Amaro Ferreira da Silva
Project 8: Problem solving environment for
Medical image analysis
The development of Medical Image Analysis (MIA) applications that can successfully be applied in clinical practice is difficult for several reasons, one of them being the large amount and variety
of resources involved (people, data, methods, computing). The lack of support for information flow from phase to phase puts extra logistics burden on the lifecycle of MIA applications. The present MSC Thesis describes efforts to develop a Problem Solving Environment (PSE) for MIA applications using the three systems available at the proof-of-concept environment of the Virtual Laboratory for e-Sciences project. The proposed PSE implements data provenance mechanisms that support information flow among systems, facilitating navigation across phases of the application lifecycle
Project 9: Streaming data between Web Services.
Comparison of streaming protocols over
a stream-enabled Web Service
The ability to stream data between web-services is vital for the implementation of complex workflows on the Grid. In this work we investigate various protocols as to their suitability for streaming, taking into consideration issues such as security, reliability and speed over different Grid configurations. To perform these comparisons we have implemented a Server/Client web service that provides a simple API through which Java applications can stream data. As an added bonus, this architecture can be reused by scientific programmers who want to stream data without having to deal with protocol or web service trivialities. Finally, an evaluation of streaming as an alternative to file transfer on Grid environments is offered, based on a number of test case scenarios
Workflow & Grid Middleware Group (WGM)