Binary system modes of matrix-coupled multidimensional Kuramoto oscillators
Chongzhi Wang,Haibin Shao,Dewei Li
2023-05-18
Abstract:Synchronization and desynchronization are the two ends on the spectrum of emergent phenomena that somehow often coexist in biological, neuronal, and physical networks. However, previous studies essentially regard their coexistence as a partition of the network units: those that are in relative synchrony and those that are not. In real-world systems, desynchrony bears subtler divisions because the interacting units are high-dimensional, like fish schooling with varying speeds in a circle, synchrony and desynchrony may occur to the different dimensions of the population. In this work, we show with an ensemble of multidimensional Kuramoto oscillators, that this property is generalizable to arbitrary dimensions: for a $d$ dimensional population, there exist $2^d$ system modes where each dimension is either synchronized or desynchronized, represented by a set of almost binary order parameters. Such phenomena are induced by a matrix coupling mechanism that goes beyond the conventional scalar-valued coupling by capturing the inter-dimensional dependence amongst multidimensional individuals, which arises naturally from physical, sociological and engineering systems. As verified by our theory, the property of the coupling matrix thoroughly affects the emergent system modes and the phase transitions toward them. By numerically demonstrating that these system modes are interchangeable through matrix manipulation, we also observe explosive synchronization/desynchronization that is induced without the conditions that are previously deemed essential. Our discovery provides theoretical analogy to the cerebral activity where the resting state and the activated state coexist unihemispherically, it also evokes a new possibility of information storage in oscillatory neural networks.
Adaptation and Self-Organizing Systems