Gaussian fitting envelope time delay estimation method and system based on generalized cross correlation

Abstract

The invention belongs to the technical field of radio orientation, radio navigation, range finding and speed measurement by use of radio waves, positioning or existence detection of reflection or re-radiation by use of the radio waves and similar devices by use of other waves and discloses a Gaussian fitting envelope time delay estimation method and system based on generalized cross correlation. According to the invention, through the generalized cross correlation method, obtaining a cross correlation function value of a signal; by use of a segmentation maximum value method, extracting envelope points of a cross correlation function; by use of Gaussian function fitting envelope, allowing the Gaussian function maximum value point to serve as a time delay estimation value of the signal. According to the invention, by use of the generalized cross correlation, cross correlation function values are calculated; by use of the segmentation maximum value envelope extraction method, envelopes ofthe cross correlation functions are extracted; and by use of the Gaussian fitting method, the envelopes are fitted; a time delay corresponding to the maximum value of the envelopes is taken as the arrival time delay difference value of two signals. The method is low in calculation complexity and easy to achieve. A simulation experiment verifies the validity and universality of the algorithm and the accuracy of the time delay estimation.

Description

technical field [0001] The invention belongs to the technical field of radio orientation; radio navigation; distance measurement or speed measurement by radio waves; positioning or presence detection by reflection or re-radiation of radio waves; similar devices using other waves, and in particular relates to a Gaussian based on generalized cross-correlation A fitting envelope delay estimation method and system. Background technique [0002] At present, the existing technologies commonly used in the industry are as follows: passive target detection occupies an important position in the fields of military affairs, national defense security, environmental monitoring, emergency rescue, military reconnaissance, etc. . Acoustic array sensor network is one of the most commonly used passive detection technologies. It shows outstanding performance in automatic detection, positioning, identification and high-speed real-time tracking, and the sound waves can bypass obstacles such as t...

AI Technical Summary

Problems solved by technology

[0004] (1) Existing technology 1 requires prior knowledge when building a model. In actual scenarios, this requirement is often difficult to meet, and the performance of the model is limited by frequency band division and deviation threshold selection. This technology does not give a reasonable optimal Frequency band division and deviation threshold selection scheme

[0005] (2) Existing technology 2 uses various signal preprocessing and noise reduction techniques to improve the accuracy of time delay estimation, but the selection, decomposition, and reconstruction of high-frequency signals in this method lead to extremely high computational complexity , which puts forward higher requirements for system storage; if the system life cycle is short, this technology will not be applicable

[0006] (3) Existing technology 3 uses wavelet noise reduction technology for signal preprocessing in the noise and reverberation environment, but the noise reduction performance of this technology is restricted by the wavelet base and decomposition scale, and the parameter is not given in this technology best solution for

[0007] (4) In order to weaken the impact of noise and reverberation on the accuracy of time delay estimation, quadratic cross-correlation, Hilbert transform, spectral analysis, matched filtering and other means are used to sharpen the GCC peak and improve the accuracy, but these technologies themselves have The performance constraints are closely related to the delay estimation accuracy, such as filter order limit, reasonable spectrum analysis high-frequency signal selection strategy determination, etc. These methods will lead to extremely high computational complexity, which will cause storage and power consumption for the system. and other burdens, and even cause the entire system to crash due to unreasonable parameter settings

[0009] In view of the above technical problems, the difficulty of solving mainly focuses on how to obtain the prior knowledge of the target sound source, the determination of the optimal setting scheme of the model parameters, and the system power consumption and memory problems caused by the optimization algorithm with high computational complexity

Images