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

1bit compression superposition CSI feedback method based on feature extraction and mutual anisotropy fusion

A feature extraction and feature extraction technology, applied in the field of superimposed feedback, can solve the problem of low CSI reconstruction accuracy, and achieve the effects of improving reconstruction efficiency, accuracy and reconstruction accuracy.

Active Publication Date: 2021-11-30
XIHUA UNIV
View PDF5 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The loss of CSI amplitude information leads to low accuracy of CSI reconstruction

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
  • 1bit compression superposition CSI feedback method based on feature extraction and mutual anisotropy fusion
  • 1bit compression superposition CSI feedback method based on feature extraction and mutual anisotropy fusion
  • 1bit compression superposition CSI feedback method based on feature extraction and mutual anisotropy fusion

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0112] In step a1), the uplink CSI estimation vector is obtained by LS estimation A specific example is as follows:

[0113] Assumption: N=2, P=4, base station receiving sequence for channel estimation for:

[0114]

[0115] The known signal s of the base station sent by the client is:

[0116] s=[0.7528-0.6083i-0.1666-0.1308i 0.9869+0.4514i 0.4556+0.2695i];

[0117] The pseudo-inverse matrix of the known signal s of the base station sent by the user end for:

[0118]

[0119] According to the LS estimation processing formula The uplink CSI estimation vector can be calculated for:

[0120]

Embodiment 2

[0122] In step a2), the length of the downlink CSI vector h compressed and quantized by the restored 1-bit compressed sensing technology is the real part of M with imaginary part and restore the feedback vector Obtain the support set of length N of the recovered downlink CSI vector h A specific example is as follows:

[0123] Assumption: N=2, M=3, restore the feedback vector for:

[0124]

[0125] According to the formula It can be obtained that the length of the downlink CSI vector h compressed and quantized by 1-bit compressed sensing technology is the real part of M for:

[0126]

[0127] The length of the recovered downlink CSI vector h compressed and quantized by 1-bit compressed sensing technology is the imaginary part of M for:

[0128]

[0129] Support set of length N of the recovered downlink CSI vector h for:

[0130]

Embodiment 3

[0132] In step a1), the vector is estimated by uplink CSI Get uplink CSI estimated vector magnitude A specific example is as follows:

[0133] The CSI estimated vector obtained in embodiment 1 according to the formula Calculate the input in the magnitude learning network for:

[0134]

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 1bit compression superposition CSI feedback method based on feature extraction and mutual anisotropy fusion, and the method comprises the steps: obtaining the learning amplitude of corresponding downlink CSI through a first neural network according to the amplitude of an uplink CSI estimation vector; performing CSI feature extraction by utilizing expert knowledge according to a recovery feedback vector obtained by the base station, and recovering a feature amplitude and a feature angle of downlink CSI; according to a downlink CSI splicing amplitude obtained by splicing the characteristic amplitude of the downlink CSI and the learning amplitude of the downlink CSI, obtaining a fusion amplitude of the downlink CSI through a second neural network; and recovering to obtain a downlink CSI reconstruction vector according to the fusion amplitude of the downlink CSI and the feature angle. Compared with single-bit CS superposition CSI feedback, the method has the advantages that the amplitude of the downlink CSI lost by the single-bit CS can be recovered according to the bidirectional heterogeneity of the uplink and downlink channels, the reconstruction precision of the CSI is greatly improved, and meanwhile, the reconstruction efficiency of the CSI is remarkably improved.

Description

technical field [0001] The present invention relates to the technical field of overlay feedback for FDD (frequency division duplex) large-scale MIMO (multiple input multiple output) systems, and in particular to a 1-bit compressed overlay channel state information (CSI, Channel State Information) feedback method based on feature extraction and heterogeneity fusion. Background technique [0002] As a key technology to meet the high spectral efficiency and energy efficiency of the future 5G (the fifth generation wireless communication) network, the FDD massive MIMO system can serve as a wireless network without increasing the transmission power and system bandwidth through hundreds of antennas deployed at the base station. More users provide wireless data services. At the same time, many performance-improving operations in the FDD massive MIMO system (such as multi-user scheduling, rate allocation, and precoding at the transmitter, etc.) depend on the acquisition of accurate d...

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): H04B7/0417H04L5/14G06N3/04G06N3/08
CPCH04B7/0417H04L5/14G06N3/08G06N3/048Y02D30/70
Inventor 卿朝进叶青刘文慧黄小莉曹太强黄永茂
Owner XIHUA 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