Parts | Days | Selection | Search | Updates | Downloads | Help

Q: Fachverband Quantenoptik und Photonik

Q 5: Quantum Optics: Open Quantum Systems

Q 5.2: Talk

Monday, March 6, 2023, 11:15–11:30, F342

Noise-induced networks — •Frederic Folz¹, Kurt Mehlhorn², and Giovanna Morigi¹ — ¹Theoretische Physik, Universität des Saarlandes, 66123 Saarbrücken, Germany — ²Algorithms and Complexity Group, Max-Planck-Institut für Informatik, Saarland Informatics Campus, 66123 Saarbrücken, Germany

We analyze a transport problem on a graph with multiple constraints and in the presence of noise and determine the network topologies to which the dynamics converges. The dynamics results from the interplay of a nonlinear interaction function and Gaussian, additive noise. The deterministic model is based on an optimization algorithm that has been designed starting from biologically-inspired models and reproduces essential elements of a neural network. The amplitude of the noise is a variable that simulates the temperature of an external bath. We show that different network topologies emerge as a function of the noise amplitude and are generally multi-stable. Remarkably, the system converges to the most robust configuration at finite noise amplitudes thereby exhibiting a resonant-like behavior. Interestingly, this configuration is not found by the deterministic dynamics and is reached with the maximal convergence. Our results suggest that stochastic dynamics can boost transport on a nonlinear network.