Network Inference using Sinusoidal Probing

@article{Delabays2020NetworkIU,
  title={Network Inference using Sinusoidal Probing},
  author={Robin Delabays and Melvyn Tyloo},
  journal={arXiv: Dynamical Systems},
  year={2020}
}
The aim of this manuscript is to present a non-invasive method to recover the network structure of a dynamical system. We propose to use a controlled probing input and to measure the response of the network, in the spirit of what is done to determine oscillation modes in large electrical networks. For a large class of dynamical systems, we show that this approach is analytically tractable and we confirm our findings by numerical simulations of networks of Kuramoto oscillators. Our approach also… Expand

Figures from this paper

References

SHOWING 1-10 OF 38 REFERENCES
Revealing network connectivity from response dynamics.
  • M. Timme
  • Computer Science, Medicine
  • Physical review letters
  • 2007
TLDR
This work considers networks of coupled phase oscillators and explicitly study their long-term stationary response to temporally constant driving, finding good predictions of the actual connectivity even for formally underdetermined problems. Expand
Spectral Identification of Networks Using Sparse Measurements
TLDR
A new method to recover global information about a network of interconnected dynamical systems based on observations made at a small number (possibly one) of its nodes is proposed, which provides efficient numerical methods to infer global information on the network from sparse local measurements at a few nodes. Expand
Estimating topology of networks.
TLDR
A method for estimating the topology of a network based on the dynamical evolution supported on the network is suggested and can be also applied when disturbances and/or modeling errors are presented. Expand
Noise bridges dynamical correlation and topology in coupled oscillator networks.
TLDR
Noise leads to a general, one-to-one correspondence between the dynamical correlation and the connections among oscillators for a variety of node dynamics and network structures, enabling an accurate prediction of the full network topology based solely on measuring the dynamicals correlation. Expand
Dynamic network structure identification with prediction error methods - basic examples
TLDR
It is shown that appropriate attention should be given to using flexible noise models, in order to allow consistent identification of the dynamics, while the use of external excitation/probing signals may reduce this need. Expand
Robustness of Synchrony in Complex Networks and Generalized Kirchhoff Indices.
TLDR
This work finds that for specific, nonaveraged perturbations, the response of synchronous states depends on the eigenvalues of the stability matrix of the unperturbed dynamics, as well as on its eigenmodes via their overlap with the perturbation vector, and introduces new graph topological indices, which are introduced as generalized Kirchhoff indices. Expand
System Size Identification from Sinusoidal Probing.
TLDR
An accurate and efficient method to determine the size of (i.e., number of agents in) a complex dynamical system by injecting a probing signal at any point of the system and measuring its response to its probing. Expand
Detecting Hidden Units and Network Size from Perceptible Dynamics
TLDR
A detection matrix is introduced that suitably arranges multiple transient time series from the subset of accessible units to detect network size via matching rank constraints, applicable across system types and interaction topologies, and applies to nonstationary dynamics near fixed points. Expand
Topological Identification in Networks of Dynamical Systems
TLDR
The paper suggests the approximation of a complex connected network with a tree in order to detect the most meaningful interconnections of a network of linear dynamical systems. Expand
Detecting hidden nodes in complex networks from time series.
  • Ri-Qi Su, Wen-Xu Wang, Y. Lai
  • Computer Science, Medicine
  • Physical review. E, Statistical, nonlinear, and soft matter physics
  • 2012
TLDR
The paradigm for detecting hidden nodes is expected to find applications in a variety of fields where identifying hidden or black-boxed objects based on a limited amount of data is of interest. Expand
...
1
2
3
4
...