Oscillations are everywhere. Neurons generate rhythmic electrical signals, the heart beats periodically, animal populations rise and fall, and power grids must maintain stable alternating currents. Such rhythms can be essential for normal function, but excessive or suppressed oscillations may also signal instability or dysfunction.
Despite their importance, predicting when a large interconnected system will shift from a steady state to an oscillatory state remains difficult. The behavior of each component depends not only on its own dynamics, but also on the number and strength of its connections, the types of interactions involved, and the time required for information or influence to travel through the network.
The researchers have now developed an analytical framework that brings these factors together. The study shows how network complexity, propagation delays and interaction types jointly determine the transition between amplitude death and sustained oscillations.
The researchers considered networks of coupled Stuart-Landau oscillators, a widely used model for studying oscillatory dynamics. They derived a relationship between effective network complexity and the critical delay at which a stable system begins to oscillate. The analysis shows that increasing network complexity generally reduces the amount of delay needed to trigger oscillations. In sufficiently complex networks, oscillations can emerge even without propagation delay.
The study further examined how different interaction types influence the transition between amplitude death and sustained oscillations. Cooperative, competitive, mixed, and random interactions were considered. The results show that the interaction type modifies the transition boundary in complexity-delay space. Cooperative interactions promote the onset of sustained oscillations at the lowest critical complexity level, whereas competitive, mixed, and random interactions require progressively higher critical complexity levels for sustained oscillations to emerge.
To examine whether the theoretical predictions remain detectable beyond ideal computer simulations, the researchers constructed a digital–analog hardware emulation platform. A microcontroller updated the network dynamics, while external electronic circuits converted selected network states into measurable voltage signals. The observed transitions were consistent with the predicted critical delays despite finite sampling, signal quantization, transistor switching and other implementation imperfections.
The agreement between theory, numerical simulations and hardware-in-the-loop emulations shows that the predicted transition is not restricted to an idealized mathematical model. Instead, the transition remains observable in a system subject to sampling, quantization, and other nonideal effects associated with practical implementations. The framework was further applied to the topology of a C. elegans neural network, illustrating its applicability to a biologically derived network.
National Science Review
Computational simulation/modeling