Infrastructure for Research towards
Ubiquitous Information Systems Project








Software for Research and Education

We have developed a new methodology for high-fidelity simulation of TCP/IP networks. The source code of our simulator based on the methodology has been released to public at

We have designed a mobile IP network, where every handheld mobile device can be a router. Together with Dartmouth, we have demonstrated such a system in the context of mobile agents.

We distribute a research compiler infrastructure called Machine SUIF. This infrastructure is part of the NSF- and DARPA-funded National Compiler Infrastructure project. Software is freely distributed via the internet at

The VINO operating system for which this grant provided equipment and system administrative support has become the system of choice for research projects in the graduate operating systems course. In the past, students have used one of the open source operating systems available (e.g., Linux, *BSD), however, the local expertise, and simplicity of our code base makes VINO an ideal platform for OS research, which was one of our initial project goals.

Deductive parsing engine, as described by Shieber, Schabes, and Pereira (1995), freely distributed at:

DeckView document browsing software, as described by Ginsburg, Marks, and Shieber (1996).

ANT virtual machine: During the course of the UI grant, we designed the ANT virtual machine that we use to teach introductory architecture and assembly language. We have recently been awarded a grant by NSF (DUE-9950239) to expand the use of ANT across the CS curriculum and for use by other Universities. The infrastructure provided by our CISE grant, made the initial development possible.

The Machine SUIF compiler for which this grant provided equipment and system administrative support has radically changed the way that we teach our graduate-level course on compilation (Computer Science 253). In the past, the course focused on the theoretical aspects of compilation and required the students to build only small pieces of a compiler in isolation. Now, we are able to teach a course that mixes theory with practice. The Optimization Programming Interface within Machine SUIF greatly reduces the start-up costs of coding compile-time analyses and transformations. The students are able to take what they learn in lecture, apply it in a state-of-the-art environment, and test it on large, realistic applications.