Pattern Selection in Multilayer Network with Adaptive Coupling

Abstract

Feed-forward effect strongly modulates collective behavior of a multiple-layer neuron network and usually facilitates synchronization as signals are propagated to deep layers. However, a full synchronization of neuron system corresponds to functional disorder. In this work, we focus on a network containing two layers as the simplest model for multiple layers to investigate pattern selection during interaction between two layers. We first confirm that the chimera state emerges in layer 1 and it also induces chimera in layer 2 when the feed-forward effect is strong enough. A cluster is discovered as a transient state which separates full synchronization and chimera state and occupy a narrow region. Second, both feed-forward and back-forward effects are considered and we discover chimera states in both layers 1 and 2 under the same parameter for a large range of parameters selection. Finally, we introduce adaptive dynamics into inter-layer rather than intra-layer couplings. Under this circumstance, chimera state can still be induced and coupling matrix will be self-organized under suitable phase parameter to guarantee chimera formation. Indeed, chimera, cluster and synchronization can propagate from one layer to another in a regular multiple network for a corresponding parameter selection. More importantly, adaptive coupling is proved to control pattern selection of neuron firing in a network and this plays a key role in encoding scheme.