A Local Search Maximum Likelihood Parameter Estimator of Chirp Signal

A local search Maximum Likelihood (ML) parameter estimator for mono-component chirp signal in low Signal-to-Noise Ratio (SNR) conditions is proposed this paper. The approach combines a deep learning denoising method with two-step estimator. denoiser utilizes residual assisted Denoising Convolutional Neural Network (DnCNN) to recover the structured component, which used denoise original observations. Following step, we employ coarse estimator, based on Time-Frequency (TF) distribution, denoised approximate estimation of parameters. Then around results, do by using ML technique achieve fine estimation. Numerical results show that outperforms several methods terms accuracy and efficiency.