High accuracy and superresolution of direction of arrival (DOA) can be achieved by means of the Hopfield's neural network (HNN). In this paper a survey for related studies of this field is presented. The HNN's model，the common steps of signal processing with the HNN and the concrete form of estimation of DOA with the HNN are described. Some methods of raising accuracy and improving resolution for it are discussed. The topics for further study are pointed out.