ISSNIP

Event Name Seminar by Gilles Gigan (James Cook University, QLD, Australia) on "The Sensor Abstraction Layer
Start Date 31st Jul 2008 4:00pm
End Date 31st Jul 2008 5:00pm
Duration 1 hour
Description

Speaker: Gilles Gigan 

Location:Brown Lecture Theatre, Dept of EEE, The University of Melbourne, Australia 

Abstract:
Traditional sensor network infrastructures rely heavily on middleware software to interface with various instruments. Each sensing technology requires its own plugin so it can be supported by the software. Many middleware solutions have been developed to suit different needs and requirements that vary from one deployment to another. At the speed at which the technology evolves, adding  extra sensors or replacing older nodes with newer ones, in an exiting network, occurs quite frequently. Existing middleware products have the following recurring problem - supporting a new technology requires programming a new plugin and loading it into the middleware software. In some instances, even re-arranging the network, by adding or removing nodes, requires a small amount of coding. In the current state, plugin-based middleware software does not offer the flexibilty required to achieve “plug & play”, where users can simply connect new sensors and instruments and expect them to work straight away. In this presentation, we introduce a Sensor Abstraction Layer (SAL) which implements all hardware-related features and enforces a clear separation of responsabilities. Using SAL, middleware software can be made hardware-independent. Hardware management functions are only implemented in SAL, which avoids code duplication and promotes reusability across middleware technologies. SAL solves many of the issues current middleware products have with respect to hardware detection and configuration, as it automates most of the tasks, making new instruments available for immediate use as soon as they are connected. A demonstration of SAL's capabilities will be made at the end of the presentation. 

Bio:
Gilles Gigan is currently employed as a Research Officer by the High Perfomance Computing center at JCU, working on projects including DART and ARCHER (DEST funded, multi institutional collaborative projects) among others. Previously, Gilles has worked as an Computer Science lecturer in the School of Maths, Physics and IT at JCU. 

top of page