Harmonics in Multiplicative and Additive Noise: Parameter Estimation Using Cyclic Statistics

Georgios B. Giannakis, Guotong Zhou

Research output: Contribution to journalArticlepeer-review

81 Scopus citations

Abstract

The problem of concern here is parameter estimation of harmonics in the presence of multiplicative and additive noise. Cyclic statistics are employed to estimate the frequencies and phases, after which the time series is demodulated and cumulants of the noise processes are estimated. The latter are then supplied to linear or nonlinear cumulant-based algorithms to identify ARMA model parameters for the noises. Cyclic statistics and higher order spectra-based approaches are shown to yield the same frequency estimates. Simulation examples illustrate the algorithms.

Original languageEnglish (US)
Pages (from-to)2217-2221
Number of pages5
JournalIEEE Transactions on Signal Processing
Volume43
Issue number9
DOIs
StatePublished - Sep 1995

Fingerprint Dive into the research topics of 'Harmonics in Multiplicative and Additive Noise: Parameter Estimation Using Cyclic Statistics'. Together they form a unique fingerprint.

Cite this