Chemical signaling as a useful metaphor for resource management

Abstract

Declarative logic programs have proved useful for resource management problems since the early 80's. However the complexity of such programs is in a direct exponential relationship with the growth in the number of resources and users. We provide a simple, easy to implement, methodology for mathematically representing requests over resources inspired by the chemical signaling model of neural networks. Our resource management model uses substructural logic in its novel incarnation, HYPROLOG, to provide a connectionist neural network representation in which requests for resources are mapped to signals triggered and consumed by resource requesters and resource consumers respectively. Through this connectionist representation model, we achieve high level of expressivity while making the model directly executable. We exemplify the power of our model through representing a concrete temporal resource scheduling information system and then apply it to some real world mathematical problems. © 2011 Springer-Verlag.