Archive Graphic

Please Note: The content on this page is not maintained after the colloquium event is completed.  As such, some links may no longer be functional.

Download Adobe PDF Reader

Matt HollandMatt Holland
Open-source Peer-to-Peer Environment to Enable Sensor Web Architecture

Wednesday, September 24, 2008
Building 3 Auditorium - 3:30 PM
(Refreshments at 3:00 PM)

A flexible, dynamic, and reliable secure peer-to-peer (P2P) communication environment is under development. Popular open-source P2P software technology provides a self-organizing, self-healing ad hoc "virtual network overlay" protocol-suite. Recent efforts have built a proof-of-concept geomagnetic Sensor Web upon this foundation. Our long-term objective is to enable an evolution of many types of distributed Earth system sensors and related processing/storage components into elements of an operational Sensor Web via integration into this P2P Environment.

In general, the Environment distributes data communication tasks among the sensors and other elements (viewed as peers, each assigned a peer- role) and controls the flow of data. This work encompasses dynamic discovery, monitoring, control, and configuration as well as autonomous operations, real-time modeling and data processing, and secure ubiquitous communications. Specifically, the recent work focuses on integrating remote geomagnetic sensors, each having operating modes to manage, with modeling processes (also with remotely managed modes). We achieved a simple form of system autonomy through a feedback loop which uses model output to drive the remote management of system elements. In addition, we have implemented basic identity management features; providing mechanisms which restrict data-serving privileges to authorized users, and which allow improved trust and accountability among users of the Environment. Browsing peers access trusted near-real-time global, and on-demand regional, representations of geomagnetic activity "nowcasted" from dynamic sensor-reported values, but also have the option to access trusted sensors directly.

Matt Holland received his Bachelor of Science in Electrical Engineering from the University of Maryland at College Park in 2001; completing a curriculum focused on computer science, signal processing and communication systems.
Originally hired while still in his senior year, his work at Goddard (mostly in Code 587 of AETD's Software Engineering Division) has been split between science data processing support of space weather and geospace missions, while collocated with the science teams, and in-house technology R&D efforts to improve Goddard's science data processing capabilities.

IS&T Colloquium Committee Host: Jacqueline LeMoigne