Unlock instant, AI-driven research and patent intelligence for your innovation.

Signal classification and identification method based on improved recurrent neural network

A cyclic neural network and signal classification technology, applied in the field of deep neural network, can solve problems such as forgetting, achieve optimization effect, improve recognition effect, and control the effect of discrimination deviation

Pending Publication Date: 2021-01-22
TAISHAN UNIV
View PDF7 Cites 4 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

As a result, RNNs forget what they saw in longer sequences and only have short-term memory

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
  • Signal classification and identification method based on improved recurrent neural network
  • Signal classification and identification method based on improved recurrent neural network
  • Signal classification and identification method based on improved recurrent neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0041]In a fully connected neural network or convolutional neural network, the network structure is from the input layer to the hidden layer to the output layer. The layers are fully or partially connected, but the nodes between each layer are not connected. This network architecture can improve the recognition and classification of various data forms by deepening the number of network layers. Meanwhile, the vanishing gradient problem is more prone to occur when the parameters are passed through the deepened network architecture. As the network structure continues to deepen, the disappearance of learned information becomes more serious when information and gradients are passed between layers of the network structure. In order to better improve the recognition effect, it is necessary to establish a connection between transfer learning and parameters between layers. This method is a good solution to the problem of gradient disappearance caused by deep network architecture.

[...

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 signal classification and identification method based on an improved recurrent neural network. The method comprises the following steps: receiving signal data; adopting a gating circulation unit as a data processing basic unit; on the basis of the GRU, adding a model average value to stack multiple layers for improvement to obtain an improved GRU; training the improved GRU as a model; and using the trained model to carry out signal data classification identification and output. The signal identification method provided by the invention not only can effectively overcome the influence of underwater acoustic communication noise, but also can resist the interference of Doppler effect in common underwater acoustic communication and efficiently identify the signal category. According to the invention, a hyper-parameter combination suitable for underwater acoustic communication is adopted to optimize the effect of a GRU identification signal, the resolution deviationis controlled, and the identification precision is improved.

Description

technical field [0001] The present invention relates to the technical field of deep neural network, in particular to a method and system of recurrent neural network based on underwater acoustic communication signal recognition. Background technique [0002] Communication signal recognition, especially the type recognition of communication signals, is the core technology of non-cooperative communication systems, and has been extensively studied in recent years. It has extraordinary value in military and civilian fields, and it has become an indispensable part in cognitive radio and software radio. In the military arena, jamming communications requires sending higher power signals to outperform the enemy's signal in the same frequency band. The key link is to have the same signal type as the enemy in order to generate higher power jamming signals. In the civilian field, the link adaptive system adaptively selects the signal type according to the channel conditions to improve...

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): G06K9/00G06K9/62G06N3/04G06N3/08
CPCG06N3/08G06N3/045G06F2218/04G06F2218/12G06F18/24
Inventor 王岩
Owner TAISHAN UNIV