Two High Accuracy Frequency Estimation Algorithms Based on New Autocorrelation-Like Function for Noncircular/Sinusoid Signal.
Kai Wang,Jiaying Ding,Yili Xia,Xu Liu,Jinguang Hao,Wenjiang Pei
DOI: https://doi.org/10.1587/transfun.e101.a.1065
2018-01-01
IEICE Transactions on Fundamentals of Electronics Communications and Computer Sciences
Abstract:Computing autocorrelation coefficient can effectively reduce the influence of additive white noise, thus estimation precision will be improved. In this paper, an autocorrelation-like function, different from the ordinary one, is defined, and is proven to own better linear predictive performance. Two algorithms for signal model are developed to achieve frequency estimates. We analyze the theoretical properties of the algorithms in the additive white Gaussian noise. The simulation results match with the theoretical values well in the sense of mean square error. The proposed algorithms compare with existing estimators, are closer to the Cramer-Rao bound (CRLB). In addition, computer simulations demonstrate that the proposed algorithms provide high accuracy and good anti-noise capability.