Accurate frequency estimation using an all-pole filter with mostly zero coefficients

An all-pole Lth order filter with only M (M < L) nonzero coefficients is used to estimate accurately the frequencies of M closely spaced sinusoidal signals in noise. The method eliminates redundancy in the linear prediction equations which caused the spurious spectral peaks observed in previously known methods [3], [4]. Since most filter coefficients are zero, the computational effort required is small.

[1]  Ramdas Kumaresan,et al.  Improved spectral resolution II , 1980, ICASSP.

[2]  Ramdas Kumaresan,et al.  Singular value decomposition and spectral analysis , 1981, 1981 20th IEEE Conference on Decision and Control including the Symposium on Adaptive Processes.

[3]  S. Lang,et al.  Frequency estimation with maximum entropy spectral estimators , 1980 .

[4]  G. W. Lank,et al.  A Semicoherent Detection and Doppler Estimation Statistic , 1973, IEEE Transactions on Aerospace and Electronic Systems.

[5]  T. Ulrych,et al.  Time series modeling and maximum entropy , 1976 .

[6]  R. Kumaresan,et al.  Estimation of frequencies of multiple sinusoids: Making linear prediction perform like maximum likelihood , 1982, Proceedings of the IEEE.