Open AccessModified Algorithm Based on Quadratic Correlation Time Delay Estimation
Author(s) -
Jun Zheng,
Yu-Zheng Zheng,
JianYu Meng
Publication year - 2022
Publication title -
international journal of computer applications technology and research
Language(s) - English
Resource type - Journals
ISSN - 2319-8656
DOI - 10.7753/ijcatr1103.1006
Subject(s) - cross correlation , algorithm , computer science , quadratic equation , reverberation , noise (video) , correlation , signal (programming language) , mathematics , acoustics , artificial intelligence , statistics , physics , geometry , image (mathematics) , programming language
In the passive positioning system, the traditional generalized cross-correlation algorithm estimates the effect under the influence of noise and reverberation, while the delay estimation algorithm after homomorphic filtering reduces the anti-reverb ability of the cross-correlation algorithm, but also reduces the cross-correlation of the signal, resulting in the decrease in the noise resistance of the cross-correlation algorithm. Based on this, the quadratic cross-correlation operation of the full-pass component of the signal is performed after homomorphic filtering to improve its noise immunity. In addition, the Introduction of Hilbert Difference sharpens the secondary cross-correlation peaks to make peak detection that reflects the time delay more accurate. Experimental simulation shows that the proposed method can effectively suppress the influence of noise and reverberation in non-stationary speech signals, and improve the performance of delay estimation.