Adaptive non-integer time delay estimation method for use in low signal-to-noise ratio impulse noise environment

Abstract

The invention provides an adaptive non-integer time delay estimation method for use in a low signal-to-noise ratio impulse noise environment. The method is characterized by comprising the following steps: performing self-covariant and cross-covariant operations on an observation sequence, and performing correlation method time delay estimation on a covariant sequence to obtain an integer bit of atime delay estimation value to serve as the initial iteration value of an LMPFTDE algorithm; then performing iteration under a minimum average p norm criterion by using the covariant sequence as an input signal of the LMPFTDE algorithm to obtain a non-integer time delay estimated value; and using the median of the iteration time delay value of a convergence process as a time delay estimated value.In the method, the observation sequence is subjected to covariant processing to weaken the influence of irrelevant noise, enhance the signal-to-noise ratio, suppress impulsive noise, and retain the correlation information between signals. After the observation sequence is subjected to covariant processing, the signal length doubles, more iterations can be carried out, and more iteration value references can make the time delay estimated value closer to a real value.

Description

technical field [0001] The invention relates to an adaptive non-integer time delay estimation method in a low signal-to-noise ratio impulse noise environment, and belongs to the technical field of adaptive processing of weak signals. Background technique [0002] Among the time delay estimation algorithms based on α-stable distribution modeling noise, the covariation algorithm, fractional low-order covariance method, and covariation correlation method can estimate the integer bits of time delay estimation. The covariation correlation method first calculates the self-covariation sequence and mutual covariation sequence of the input signal. The covariation operation can suppress the impulse noise, improve the signal-to-noise ratio, and at the same time retain the phase information between the original signals. , the estimation accuracy of covariance correlation method is higher than that of covariance method and correlation method. [0003] There are two ways to improve the r...

AI Technical Summary

Problems solved by technology

[0004] For time delay estimation of some low-frequency signals, if the sampling frequency only satisfies the sampling theorem, the error of fractional bits will have a great impact on positioning; increasing the sampling rate of the signal can improve the resolution of time delay estimation, but the correlation peak is not sharp and difficult to find peak position

Images