Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Firm underdetermined blind separation source number and hybrid matrix estimating method and device

An underdetermined blind separation and mixed matrix technology, which is applied in speech analysis, instruments, etc., can solve problems such as the direction offset of the mixed matrix column vector, blind separation failure, narrow application range, etc., to reduce the amount of calculation and improve timeliness , a wide range of effects

Inactive Publication Date: 2016-02-24
TIANJIN UNIV
View PDF5 Cites 9 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] 2) The requirements on the distribution of source signal components are too strict, and the application range is narrow
When there are overlapping frequencies, it can lead to incorrect estimates of the number of sources or a shift in the direction of the column vectors of the mixing matrix
eventually lead to blind separation failure

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Firm underdetermined blind separation source number and hybrid matrix estimating method and device
  • Firm underdetermined blind separation source number and hybrid matrix estimating method and device
  • Firm underdetermined blind separation source number and hybrid matrix estimating method and device

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0063] In order to reduce the computational complexity of blind separation and expand the application range of blind signal separation, the embodiment of the present invention proposes a robust underdetermined blind separation source number and mixing matrix estimation method, which is aimed at periodic or approximately periodic stationary signals, Different source signals are allowed to have frequency overlap in a certain range, which reduces the requirement for source signal sparsity. The method includes the following steps:

[0064] 101: Perform spectrum correction on the observed spectrum; use the spectrum correction result to construct an M×Q dimensional harmonic parameter matrix, and normalize the constructed matrix;

[0065] 102: Establish a distance matrix for the normalized matrix, judge whether the elements in each column of the distance matrix are less than the threshold ξ, the frequency corresponding to the column whose elements are not less than the threshold is th...

Embodiment 2

[0069] The scheme in embodiment 1 is described in detail below in conjunction with specific calculation formula, accompanying drawing, example, see below for details:

[0070] According to different models of hybrid systems, blind separation can be divided into linear instantaneous hybrid, linear convolutional hybrid, and nonlinear hybrid. Linear transient mixing is the simplest structural form in blind signal separation, and it is also the most classic mixing model form. Therefore, the embodiment of the present invention adopts the linear transient mixing system model, and its model is:

[0071] x(t)=As(t)+n(t)(1)

[0072] Among them, s(t)=[s 1 (t),s 2 (t)...s n (t)...s N (t)] T is the source signal vector, N is the number of source signals; x(t)=[x 1 (t),x 2 (t)...x m (t)...x M (t)] T is the observed signal vector, M is the number of observed signals; n(t)=[n 1 (t),n 2 (t)...n n (t)...n N (t)] T is the noise, and A is the M×N dimensional unknown mixing matrix....

example 1

[0077] Example 1: Suppose the number of sources N=3, the number of observations M=2, and the sampling rate f s =8000Hz, the number of sampling points is L=4096, so that the direction of the mixing matrix is ​​20°, 80°, 140°, then the mixing matrix A is

[0078]

[0079] The N=3 source signals contain a total of Q=10 frequencies. Table 1 shows the amplitude and phase parameters of the frequency components of each source signal. The expression of each source signal is

[0080]

[0081] Among them, for the qth frequency component (1≤q≤Q)f q , a n,q and respectively for the nth source signal s n The amplitude and phase of the corresponding harmonic components in (t).

[0082] Table 1 Amplitude and phase parameters of the frequency components of each source signal

[0083]

[0084]

[0085] It can be seen from Table 1 that overall, s 1 (t)~s 3 The frequency component distribution of (t) is sparse, but not strictly sparse. Specifically, s 1 (t) and s 2 (t) Th...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a firm underdetermined blind separation source number and hybrid matrix estimating method and device. The method comprises the following steps: correcting an observed frequency spectrum; constructing an M*Q-dimension harmonic parameter matrix by using a spectrum correcting result and carrying out normalization on the constructed matrix; establishing a distance matrix for the normalized matrix; judging whether elements in different lines of the distance matrix are smaller than a threshold value ksai or not, defining the frequency corresponding to lists with elements not smaller than the threshold value as overlapped frequency, and rejecting the lists corresponding to the frequency from the normalized matrix; calculating a potential function of the processed matrix; searching a peak value of the potential function; and determining source number estimation and hybrid matrix estimation. The device comprises an inputting module, a processing module and an outputting module. By the firm underdetermined blind separation source number and hybrid matrix estimating method and device, source signal number and a hybrid matrix of an underdetermined blind separation system for cyclic stationary signals can be estimated precisely; and observation signals are processed by a frequency spectrum correcting method, and the potential function can be gained quickly, follow-up calculated amount is reduced, and timeliness is improved.

Description

technical field [0001] The present invention relates to the technical field of digital signal processing, in particular to a method and device for robust underdetermined blind separation of source numbers and mixing matrix estimation, and in particular to a method and device for the case where the number of periodic observation signals is smaller than the number of source signals when there is frequency overlap between source signals In the case of blind signal separation (that is, the underdetermined case), the number of sources and the mixing matrix are estimated only by processing the observed signals. Background technique [0002] Blind Source Separation (BSS) refers to the process of recovering the source signal by only using the observation signal of the sensor when the source signal and the mixed channel are unknown. [1] . This problem is widely used in the field of signal processing. According to the relationship between the number of observations M and the number ...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
IPC IPC(8): G10L21/0272
CPCG10L21/0272
Inventor 黄翔东靳旭康
Owner TIANJIN UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products