Signal Spectrum Estimation of Parameter Model Approach Based on MATLAB

Parameter model methods of power spectrum estimation in electronic signal processing field have been introduced including Levinson-Durbin Method, Burg Method and the latter proposed Covariance Method. The measure of the random processing signal spectrum has been investigated based on the mathematic software MATLAB. The performance of spectrum estimation such as the analyzing of signal frequency, resolution, and side lobe characters has been realized in the research. Different parameter values of method has been compared, conclusions can be made that larger the sampling number is, higher estimation performance will be, and Covariance Method is equipped with a satisfying performance to its counterparts although the appearance of frequency fluctuation in side band.