Global exponential stability of cellular neural networks with time-varying delays

The existence of equilibrium point and global exponential stability (GES) for cellular neural networks with time-varying delay are explored in this paper by applying the extended Halanay's delay differential inequality, the theory of homotopy invariance, Dini's derivative, and several functional analysis techniques. Some simple and new sufficient conditions are obtained to ensure existence, uniqueness of the equilibrium point and its GES of the neural networks. The results are less conservative than those established in the previous literature. In addition, this condition requires neither the active functions to be differentiable, bounded, and monotone nondecreasing nor the time-varying delays to be differentiable.