Explosive death transitions in complex networks of limit cycle and chaotic systems

The conditions for explosive death transitions in complex networks of oscillators having generalized network topology, coupled via mean-field diffusion, are derived analytically. The network behaviour is characterized using three order parameters that define the average amplitude of oscillations, the mean state and the fraction of dead oscillators. As the mean field coupling is changed adiabatically, the amplitude order parameter undergoes explosive death transitions and the nodes in the network collectively cease to oscillate. The transition points in the parameter space and the boundaries of amplitude death regime are derived analytically. Sub-critical Hopf bifurcations are shown to be responsible for amplitude death transition. Explosive death transitions are characterized by hysteresis on adiabatic change of the bifurcation parameter and surprisingly, the backward transition point is shown to be independent of network topology. The theoretical developments have been validated through numerical examples, involving networks of limit cycle systems — such as the Van der Pol oscillator and chaotic systems, such as the Rossler attractor for both random and small world networks. •Analytical conditions for explosive death transitions are derived.•Sub-critical Hopf bifurcations responsible for explosive death.•Explosive death transitions exhibit hysteresis on adiabatic change in parameters.•Forward death transition dependent on network topology.•Backward death transition independent of network topology.