Thermo-inspired modeling and analysis of network information flows

The aim of this research is to develop a fundamental thermo-inspired framework for modeling and analysis of network information flows. This framework is partially motivated by the treatment of energy flows in thermodynamic systems and Eulerian modeling of continuum network information flows as well as non-local biological aggregations. Capturing intrinsic energy or information flows inside these network systems is a crucial problem, partly because the energy or information flows within the network systems are the possible way to uncover the essence of stability and other dynamic properties of the network system and will definitely help us understand some fundamental phenomena exhibited by the natural and engineered systems. Furthermore, the existing theories for information flow modeling hardly address any systematic synthesis methods for building engineered complex systems, which may revolutionize the control theory and applications to network analysis and synthesis as well as diffusion processes. The proposed modeling and analysis framework in this paper is based on our recent research related to system thermodynamic theory in which energy flow is the central part of this theory and is consistent with some basic thermodynamic properties such as energy conservation and entropy nonconservation. More specifically, the proposed framework is inspired by the recently developed notion of system thermodynamics which results in model architectures involving the information flow propagated over a phase space according to certain thermodynamic laws.