Return to d'Auriol's Seminar Page
Brian J. d'Auriol, Ubiquitous Sensor Network Visualization Models, Invited Colloquium talk, April 18, 2007, Department of Computer Science, National Tsing Hua University, Taiwan.
Ubiquitous Sensor Networks (USNs) have the potential to generate large amounts of time
sequenced raw data anytime and anywhere. In addition, the USN itself forms a dynamic system. As
such, information about the system including its state variables and state transitions
increases the data requirements inherent in USNs. Moreover, large-scale USNs including
clustered organization, routing algorithms and power constraints increase the complexity of the
system, further adding to the information requirements. In many cases, it is important for
humans to understand the implications of the raw data or of the underlying system that the raw
data is associated with, or, the operation of the USN itself. This is especially the case when
supporting policy-based decision-making. Visualization is well-known to facilitate human
understanding of data and its underlying systems. This talk proposes several visualization
models including the Parameterized Iconic Glyph (PIG) and the Orthogonal Organized Finite State
Machine (OOFSM) models as potential candidates for visualization models capable of addressing
the large data size issues inherent in USNs. Example applications of these models to sensor
acquired information are also described.