Effects of clustering on diversity-induced resonance in hidden metric spaces

The effects of network clustering on diversity-induced resonance of an ensemble of bistable systems subjected to weak signals are investigated. A class of network models based on hidden metric spaces is used to generate networks with different levels of clustering. The propensity of triples of nodes is to form triangles. By numerical simulations and analytical calculations, it is shown that the maximum response is suppressed by network clustering. The optimal diversity strength for maximum response is also reduced. A pronounced resonance phenomenon is observed at an intermediate clustering strength. Either low or high clustering strength will undermine the response of the systems to external signals. The results imply that it is possible to control the response of spatially extended systems to external signals by the manipulation of network clustering.