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

Censored data parameter self-adaption estimation method based on information theory learning

A technology of adaptive estimation and learning method, applied in the direction of baseband system components, etc., can solve problems such as satisfying Gaussian distribution

Inactive Publication Date: 2015-09-09
HANGZHOU DIANZI UNIV
View PDF4 Cites 3 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, in many cases the measurement noise n k,i does not necessarily satisfy the Gaussian distribution

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
  • Censored data parameter self-adaption estimation method based on information theory learning
  • Censored data parameter self-adaption estimation method based on information theory learning
  • Censored data parameter self-adaption estimation method based on information theory learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0081] Suppose a sensor network contains K=20 nodes, and the vector w to be estimated 0 is a normalized 5×1 vector (ie ||w 0 || 2 = 1). noise n k,i It is generalized Gaussian noise, that is, the probability density of the noise satisfies f(n)∝exp(-|v| p ), where p is the shape parameter, when 0k,i is super-Gaussian noise (in particular, p=1, n k,i is super Laplacian noise), when p=2, n k,i is Gaussian noise, when p>2, n k,i is sub-Gaussian noise. Define the signal-to-noise ratio as: In addition, the width parameter adopted by the Gaussian kernel function in this method is σ=2. In the following experiments, the present method is compared with MSE-based adaptive methods:

[0082] Experiment 1: In the case of SNR=5dB, calculate the relationship between the mean square estimation error and the iterative cycle i, the results are as attached image 3 It is shown and shown that no matter in super-Gaussian, Gaussian or sub-Gaussian noise environment, the method of the prese...

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 censored data parameter self-adaption estimation method based on information theory learning. The invention comprises following steps of 1: designing estimation error functions of censored signals; 2: calculating secondary Renyi entropy of the estimation errors; 3: obtaining local estimation of all node parameters by means of gradient descent; 4: designing weighing coefficient according to local estimation of each node and neighbor node thereof; and 5: exchanging local estimation of the node and the neighbor node and calculating weighing estimation of the node and the neighbor node thereof by use of the weighing coefficient obtained in the step 4. According to the invention, 1) self-adaption estimation of parameters included in the censored data is achieved; 2) collaborative estimation of all nodes of a sensor network is achieved; 3) impacts on estimation performance by signal abnormal values caused by attacks suffered by the network nodes can be effectively reduced; and 4) higher estimation precision can be provided under a gaussian or non-gausssian noise environment.

Description

technical field [0001] The invention belongs to the technical field of statistical signal processing, in particular to an adaptive estimation method of censored data parameters based on information theory learning. Background technique [0002] A sensor network is a distributed network system composed of a large number of sensor nodes, which can collaboratively monitor, perceive and collect physical information of various environments or monitoring objects in the network coverage area in real time, and process and transmit it. Signal parameter estimation is an important application of sensor network. It uses various algorithms to obtain estimates of unknown physical quantities (such as temperature, target orientation, motion speed, etc.) from the measurement values ​​of multiple sensors polluted by noise. The distributed estimation method does not require a central processing unit, and has high reliability and robustness; it reduces the data transmission and processing of th...

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): H04L25/02
Inventor 刘兆霆余旺科
Owner HANGZHOU DIANZI 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

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