University of Minnesota
Sensor Network Studio
index.php

Overview

This is a joint project with three participating universities: University of Illinois at Urbana-Champaign, University of Virginia and University of Minnesota.  We are developing a new development environment with support for multiple phases of the sensor network software life-cycle. It includes tools for sensor network design, implementation, execution, and post-deployment troubleshooting and maintenance. The goal is to create usable sensor network development support for the community and reduce sensor network adoption barriers for a broader category of users.

The proposal complements earlier work in three main respects: 1) Design-Time Support Operating System,  2) Programming and Network Architecture Support and 3) In-Situ System Support

Research Directions

 

 we propose to develop a light-weight object oriented operating system, called LiteOS, to enable object-oriented programming for sensor networks. LiteOS will provide object-based encapsulation of local operating system resources. It will export a set of C++ class libraries for access to resources such as flash, radio, and timers. C++ is chosen because of its widely known syntax and its support for objects.
This project aims at an Architecture for Worldwide Cyber-Phyical Computing.  It is  a multi-layer architecture for interconnecting and programming heterogeneous worldwide sensor nets, which defines clear interfaces for each layer.  Specifically,  higher layers export high-level abstractions that are easy to use and lower layers export low-level abstractions to provide full flexibility

In this project, we design and implement a practical Sensing Area Modeling technique, called SAM. By injecting controlled events through regular and hierarchical training, SAM estimates the sensing areas of individual sensor nodes accurately. We also propose several model abstraction techniques to concisely define the sensing areas with a small loss of accuracy.  This work is the first to investigate the impact of irregular sensing area on application performance, such as coverage scheduling and tracking. We evaluate SAM using theoretical analysis, a physical experiment on a testbed consisting of 40 MicaZ motes, as well as an extensive 1000- node simulation. Our evaluation results reveal serious problems caused by circular sensing model, while demonstrating significant performance improvements when SAM is used.

We are developing a design tool ANDES (ANalysis-based DEsign tool for wireless Sensor networks) for modeling and analyzing a wireless sensor network (WSN) system before deployment. This is beneficial for the system designers who can develop a system model and refine it iteratively, by tuning the system parameters based on existing analytical results, and decide on the final system configuration
according to the desired system performance. This is one of the first works being done to integrate analysis techniques into a design tool for WSNs. For implementing ANDES, we have extended the AADL/OSATE framework, which has been used extensively for real-time
and embedded systems.

This Page was last modified by 11/21/2006

Authors: Tian He 

 

The views and opinions expressed in this page are strictly those of the page author.
The contents of this page have not been reviewed or approved by the University of Minnesota.