Anti-interference method for communication based on deep deterministic gradient reinforced learning

A reinforcement learning and gradient technology, applied in neural learning methods, interference to communication, biological neural network models, etc., can solve problems such as quantization errors and power selection results that cannot be optimal, to overcome quantization errors and reduce network complexity , the effect of improving performance

Inactive Publication Date: 2019-02-01
UNIV OF ELECTRONICS SCI & TECH OF CHINA +1
View PDF3 Cites 34 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, this method will introduce quantization error when quantizing an...

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
  • Anti-interference method for communication based on deep deterministic gradient reinforced learning
  • Anti-interference method for communication based on deep deterministic gradient reinforced learning
  • Anti-interference method for communication based on deep deterministic gradient reinforced learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0036] figure 1 It is the specific implementation method of the algorithm of the present invention, combined below figure 1 Describe each step and its principle in detail.

[0037] The algorithm implementation framework of the method proposed by the present invention based on the depth-determined gradient strategy reinforcement learning continuous strategy selection anti-interference method is as follows: figure 1 shown. In step S1, S1.1 completes interference and wireless environment modeling. In the scenario, multiple interference sources interfere with legal communication links. The interference methods may include but are not limited to: single-tone interference, multi-tone interference, linear frequency sweep interference, partial frequency band interference, and noise frequency hopping interference. The interference source can realize the dynamic adjustment of interference to legitimate users by adjusting interference parameters or switching interference modes. The s...

Embodiment 2

[0083] The convolutional neural network structure proposed by the present invention for anti-interference decision-making is as follows: figure 2 As shown: In the simulation, it is assumed that the system is divided into 128 sub-channels, and a 128×128 spectrum time slot state matrix is ​​constructed according to the spectrum sampling signal as the input of the convolutional neural network; then three convolutional layers, two pooling layers and two The fully connected layer outputs a 1×128 power vector. Specifically, the convolutional layer, pooling layer and operations in the convolutional neural network are as follows:

[0084] Assuming that the input data of the convolution operation is I, the corresponding convolution kernel K has the same dimension as the input data. Take three-dimensional input data as an example (when the input data is two-dimensional, the third dimension can be regarded as 1). The convolution operation requires the third dimension of the convolutio...

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 belongs to the technical field of wireless communication, and relates to an anti-interference method for communication based on deep deterministic gradient reinforced learning. The method provided by the invention comprises the steps of firstly building an interference environment model according to the quantity of interference sources and a wireless channel model; building a utilityfunction according to a legal user communication quality index, using the utility function as a return in learning; forming spectrum information sampled at different time slots into a spectrum time slot matrix, and describing an interference environment state by using the matrix; and then building a convolutional neural network according to a deep deterministic gradient reinforced learning mechanism, and when an anti-interference decision is made, an environment state matrix achieves anti-interference strategy selection of the corresponding state in a continuous space via a target actor convolutional neural network. According to the method provided by the invention, the continuous anti-interference strategy selection in communication is completed based on the deep deterministic gradient strategy reinforced learning mechanism. The quantization error caused by quantized discrete processing on the policy space is overcome, the quantity of cells output by the neural network and the network complexity are reduced, and the anti-interference algorithm performance is improved.

Description

technical field [0001] The invention belongs to the technical field of wireless communication, and relates to a communication anti-jamming method based on depth determination strategy gradient reinforcement learning. Background technique [0002] With the development of wireless communication technology, the electromagnetic environment faced by wireless communication systems is increasingly complex and harsh. It may suffer from unintentional interference from its own communication, and may also be affected by interference signals intentionally released by the enemy. The traditional anti-jamming means adopt fixed anti-jamming strategies for the static jamming mode of the jamming source. With the intelligentization of jamming methods, jamming sources can dynamically adjust jamming strategies according to changes in the communication status of legitimate users, making traditional anti-jamming methods unable to guarantee normal communication of legitimate users in a dynamic jamm...

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): H04K3/00G06N3/04G06N3/08
CPCH04K3/20H04K3/40G06N3/08G06N3/045
Inventor 黎伟王军李黎党泽王杨
Owner UNIV OF ELECTRONICS SCI & TECH OF CHINA
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap