Perceiving agents have a limited sensitivity to distinguish input stimuli.
Whatever the cause-random noise, limited numbers of sensing elements, inten
tional desensitization-formulation of such behavioral phenomena at the syst
em level has not attracted much interest. This work is a first attempt to f
ormalize the limited-sensitivity abstraction and to explore its implication
s for systems modeling and simulation, both conceptual and practical. In th
is article. a system theoretic formalization of perceiving agents called qu
antized systems is provided, linking them with discrete event modeling para
digms and exploring their approximation capability and message reduction in
distributed simulation.