Design of active network filters as hysteretic sensors

This work aims to propose and design a class of networks coupled linear nonlinear oscillators, in which short bursts exogenous excitation result sustained endogenous network activity that returns quiescent state only after characteristic time along different path than when originally excited. The desired hysteretic behavior is obtained through the coupling self-excited oscillations with purposely designed rate laws for slowly varying nodal parameters, governed by local interactions network. proposed architecture sought dynamics take inspiration from complex biological systems combine energy sources paradigm distributed sensing information processing. In this paper, problem considers arbitrary topologies investigates dependence response on model as well placement single node an otherwise Perturbation analysis various asymptotic parameter limits used define internal dynamics. Parameter continuation techniques validate results numerically demonstrate their robustness over finite ranges values. Both approaches suggest nontrivial optimal distribution nonlinearity topology.