Abstract: This paper presents a neural network based parameter estimation method to solve the problem of parameter estimation of useful signals in chaotic noise. This algorithm combines the ability of neural network approximating any nonlinear function and the spectral analysis technique together. The phase space reconstruction technique is introduced, and the principle and steps of the proposed algorithm are given out. Principle analysis and simulation experiment results of the proposed algorithm in actual application are also given out, which show the effectiveness of the algorithm for parameter estimation.