Existence and global attractivity of almost periodic solution for DCNNs with time-varying coefficients

By using exponential dichotomy, the Banach fixed point theory and some inequality analysis technology, some sufficient conditions are derived ensuring existence, uniqueness and global attractivity of almost periodic solution of delayed cellular neural networks (DCNNs) with time-varying coefficients. Without assuming the boundedness of signal functions, the results obtained have sufficient significance in design and applications of almost periodic oscillatory DCNNs.