Title :
An ARMA model for power spectrum estimation
Author :
Talkhan, E.A. ; Hassan, A.F. ; Amin, N.A.
Author_Institution :
Cairo University, Giza, Egypt
Abstract :
A new ARMA model is proposed to get the PSDF of noisy random ergodic zero mean discrete time signals. In this model, the residual power which is not covered by the AR polynomial is represented by a limited order MA polynomial. A new algorithm is proposed and used in minimizing the residual power which is still not represented by the added limited MA polynomial. A comparison is made between the proposed model and that of Kay [9] which showed that for the same degree of performance, the new model needs much less storage space and computation time.
Keywords :
Autoregressive processes; Density functional theory; Gaussian noise; Parameter estimation; Polynomials; Power engineering and energy; Signal detection; Signal resolution; Spectral analysis; Wideband;
Conference_Titel :
Acoustics, Speech, and Signal Processing, IEEE International Conference on ICASSP '83.
DOI :
10.1109/ICASSP.1983.1171974