Abstract

To optimize the performance of the existing algorithm for autoregressive (AR) model spectral estimation, an improved algorithm is presented based on linear prediction theory and Cholesky decomposition in this paper. Firstly, the principles of the proposed algorithm are described in detailed. Then, the simulation demonstrates the reliability and efficiency of the algorithm. Compared with the traditional algorithms, the improved algorithm for AR model spectral estimation effectively reduces the computational complexity with high resolution.

Paper Details