Global exponential stability and periodicity of cellular neural networks with variable delays

The Letter presents sufficient conditions ensuring the global exponential stability and existence of the periodic solution for cellular neural networks with variable delays. The results allow for the consideration of all unbounded neuron activation functions (but not necessarily surjective), in particular, can analyze the exponential stability and periodicity for the linear cellular neural networks. The work provides one such method which can be applied to cellular neural networks systems with variable delays. The method, based on the theory of fixed point and differential inequality technique. The applicability of the present results is demonstrated by two examples