ESTIMATION OF PARAMETERS OF EXPONENTIALLY DAMPED SINUSOIDS USING FASTMAXIMUM-LIKELIHOOD-ESTIMATION WITH APPLICATION TO NMR-SPECTROSCOPY DATA

In this paper, we present fast maximum likelihood (FML) estimation of parameters of multiple exponentially damped sinusoids, The FML algorit hm was motivated by the desire to analyze data that have many closely spaced components, such as the NMR spectroscopy data of human blood pl asma, The computational efficiency of FML lies in reducing the multidi mensional search involved in ML estimation into multiple 1-D searches, This is achieved by using our knowledge of the shape of the compresse d likelihood function (CLF) in the parameter space. The proposed FML a lgorithm is an iterative method that decomposes the original data into its constituent signal components and estimates the parameters of the individual components efficiently using our knowledge of the shape of the CLF, The other striking features of the proposed algorithm are th at it provides procedures for initialization, has a fast converging it eration stage, and makes use of the information extracted in prelimina ry iterations to segment the data suitably to increase the effective s ignal-to-noise ratio (SNR). The computational complexity and the perfo rmance of the proposed algorithm are compared with other existing meth ods such as those based on linear prediction, KiSS/IQML, alternating p rojections (AP), and (EM).