>



We will be presenting at the 2006 ACM conference in Melbourne Florida. Our poster presentation can be found here: Poster Presentation for ACM Conference

Design, Implementation, and Optimization of Features in a Mobile Computational Language for Internet Programming

This project focuses on mobile computations for sharing resources on the Internet. A mobile computation is a program that can commence execution at one site, discover a need for a resource at a different site, halt execution, migrate, and continue execution at the new site. Migrating an executing computation is a non-trivial issue that is exacerbated by the heterogeneity of the computing infrastructure. Native support is an essential prerequisite for a programming language that can produce efficient and secure strongly mobile computations. This project will build upon an existing prototype Mobile Computational Model, by adding functionality and optimizing the performance. A major contribution of this work will be to bridge the gap between the grid middleware and the application interface by providing a programming language and virtual machine for efficient and secure strongly mobile applications.

Research Questions

  • Will the addition of functionality to support data structures like multi-dimensional arrays adversely affect the performance of the RVM?

    Language support for data structures, like multi-dimensional arrays, is essential for building reasonable applications. Memory management and access during execution become important issues to be addressed in terms of execution efficiency, and in the context of closure capture. Click here for more information.

  • Will communication between computations adversely affect the performance of the RVM?

    Communication between computations is an essential feature of grid applications, but is complicated by the uncertain and changing locations of the computations. Click here for more information.

  • How can the RVM be optimized to further improve efficiency?

    Existing techniques for optimization of compilers and programs cannot be adapted directly due to necessity of capturing closure and restoring it at the destination host, which may have a different architecture than the originating host. Click here for more information.

This project was funded in part by a research award through the Collaborative Research Environment for Undergraduates in Computer Science and Engineering (CREU) program from the Computing Research Association Committee on the Status of Women in Computing Research (CRA-W).

Homepage of CRA-W

Homepage of The College of New Jersey