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

Feedback blind equalization method of dynamic wavelet neural network based on fuzzy control

A technology of fuzzy neural network and dynamic wavelet, which is applied to the shaping network in the transmitter/receiver, baseband system components, etc., can solve the problems of not adapting to time-varying characteristics, slow learning convergence speed, and sensitivity to external noise

Inactive Publication Date: 2010-12-01
NANJING UNIV OF INFORMATION SCI & TECH
View PDF1 Cites 14 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Intelligent Control Theory and Technology [M]. Beijing: Tsinghua University Press, 1997), because the static multi-layer feed-forward network is used to perform blind equalization on the dynamic process, many problems will inevitably arise: such as not having the function of adapting to time-varying characteristics, etc. ; Especially with the increase of the system order, the rapidly expanding network structure will make the learning convergence speed slower; in addition, more input nodes will also make the system particularly sensitive to external noise. For the blind equalization of the dynamic process, the present invention The dynamic wavelet neural network feedback blind equalization method of fuzzy control (FDWNN, Dynamic Wavelet Neural Networks based on Fuzzy controlling) provides an effective method

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
  • Feedback blind equalization method of dynamic wavelet neural network based on fuzzy control
  • Feedback blind equalization method of dynamic wavelet neural network based on fuzzy control
  • Feedback blind equalization method of dynamic wavelet neural network based on fuzzy control

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0177] Example 1: Minimum Phase Underwater Acoustic Channel Simulation

[0178] The minimum phase underwater acoustic channel, its transfer function is:

[0179] c=[0.9656 -0.0906 0.0578 0.2368]

[0180] The transmitted signal is 4QAM, the signal-to-noise ratio is 20dB, the superimposed noise is Gaussian white noise, the WNN equalizer, the weight length is 16, and the eighth tap is initialized to 1. In formula (8), C WNN =10, λ in formula (14) WNN =0.46, the initialization of wavelet function scale factor and translation factor are respectively a WNN =4.3,b WNN =0.0025; For FDWNN equalizer, the weight length is 16, the 8th tap is 1, the weight length of the transversal filter is 16, 1 / 4 tap is used, the step size μ=0.001, the initialization of wavelet function scale factor and translation factor respectively a FDWNN =3,b FDWNN =0.006, the two weighting factors in the formula (12) are respectively α=0.355, β=0.645, C in the formula (8) FDWNN =1, λ in formula (14) FDWNN ...

Embodiment 2

[0182] Embodiment 2: Sparse underwater acoustic channel simulation

[0183] The transfer function of the sparse underwater acoustic channel used in this experiment is

[0184] c=zeros(1,1001); c(1)=0.076; c(2)=0.122; c(1001)=1

[0185]The transmitted signal is 8PSK, the signal-to-noise ratio is 20dB, the superimposed noise is Gaussian white noise, WNN equalizer, the weight length is 16, the sixth tap is initialized to 1, C in formula (8) WNN =0.28, λ in formula (14) WNN =5.3, the initialization of wavelet function scale factor and translation factor are respectively a WNN =5.2,b WNN =1.685; FDWNN equalizer, weight length is 16, adopts 1 / 2 tap, the weight length of transversal filter is 16, adopts 1 / 4 tap, step size μ=0.0001, the initialization of wavelet function scale factor and translation factor are respectively a FDWNN =4.35,b FDWNN =0.007, the two weighting factors in the formula (12) are respectively α=0.9075, β=0.0925, C in the formula (8) FDWNN =0.0001, λ in for...

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 feedback blind equalization method of a dynamic wavelet neural network based on fuzzy control, which comprises the following steps: a) passing a transmitted signal sequence x(n) through an unknown channel h(n), and then superimposing on a Gaussian white noise N(n) to obtain an observation sequence y(n); b) processing an error signal e(n) by a constant modular algorithm (CMA) to obtain a tap coefficient c(n) of a linear segment formed by transversal filters in a wavelet neural network; c) passing an input value deviation E(n) and a deviation change deltaE(n) of a fuzzy neural network controller through a fuzzy neural network controller to obtain an iteration step change value delta mu of an extension factor and a shift factor of a wavelet function in a nonlinear segment formed by wavelet neural networks in a dynamic wavelet neural network; and d) sequentially passing the observation sequence y(n) through the dynamic wavelet neural network and a judger to obtain an output signal. The method has faster convergence speed and smaller steady-state error, thereby being completely applicable to underwater acoustic channels.

Description

technical field [0001] The invention relates to a dynamic wavelet neural network feedback blind equalization method, in particular to a fuzzy control dynamic wavelet neural network feedback blind equalization method. Background technique [0002] In wireless and digital communication systems, due to the influence of reflection, diffusion and scattering, there will be more complex propagation mechanisms, such as multipath effects, shadow effects and fading effects, which cause the channel to change with the user's position and time The received signal power will also fluctuate rapidly, resulting in inter-symbol interference (ISI: Inter-Symbol Interference). In order to eliminate inter-symbol interference, a blind equalizer that can adaptively equalize, adjust parameters, and track channel characteristics is generally introduced at the receiving end to complete the best estimation of the signal. Interference (see literature: Xiao Ying. Research on blind equalization algorithm...

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
Patent Type & Authority Applications(China)
IPC IPC(8): H04L25/03H04L25/02
Inventor 郭业才王丽华
Owner NANJING UNIV OF INFORMATION SCI & TECH
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

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com