Stochastic control of spiking activity bump expansion: monotonic and resonant phenomena

Abstract

We consider spatially localized spiking activity patterns, so-called bumps, in ensembles of bistable spiking oscillators. The bistability consists in the coexistence of self-sustained spiking dynamics and quiescent steady-state regime. We show numerically that the processes of growth or contraction of such patterns can be controlled by varying the intensity of multiplicative noise. In particular, the effect of the noise is monotonic in an ensemble of the coupled Hindmarsh-Rose oscillators. On the other hand, in another model proposed by V. Semenov et al. in 2016 (see Ref. [V. Semenov et al., Phys. Rev. E 93, 052210 (2016)]), a resonant noise effect is observed. In that model, stabilization of the activity bump expansion is achieved at an appropriate noise level, and the noise effect reverses with a further increase in noise intensity. Moreover, we show the constructive role of nonlocal coupling which allows to save domains and fronts being totally destroyed due to the action of noise in the case of local coupling.

5 pages, 3 figures

References in corpus (6)

• Stochastic population dynamics in spatially extended predator-prey systems
• Coherence resonance in a network of FitzHugh-Nagumo systems: interplay of noise, time-delay and topology
• Bumps and Oscillons in Networks of Spiking Neurons
• Control of coherence resonance in multiplex neural networks
• Multiplexing-based control of stochastic resonance
• Ising model with memory: coarsening and persistence properties