Are there thermodynamic variables for networked sensor systems?
John Edward Gray, Amy S. Smith-Carroll · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 2003
Discussed is a number of mechanisms for complex systems that can lead to unexpected behavior. All are related to graph theoretic models of interactions. In addition, the theory of random interval graphs can be applied to the characterization of simultaneously occurring variable finite length events. The theory is sufficiently general to provide simple parametric description of the natural relationships between a system's task processing rate and associated decision and communication rates required to control it. One can estimate conditions which induce temporal decoupling and provide general insight into control methodologies which can be used to avoid decoupling. The usage of intervals to characterize resource allocation problems has widespread applications to processing and decision making algorithms which consumed finite bounded time intervals. Finally, various aspects of track quality are discussed as a measure of the strength of interaction.