Communication signal modulation mode identification method based on convolutional neural network

Abstract

The invention discloses a modulation mode identification system and method based on a convolutional neural network, which solve the problems of complex feature extraction steps and low identificationrate under a low signal-to-noise ratio in the prior art. The simple feature in the identification system is constructed as a simple feature using a co-directional component and a quadrature componentof a baseband signal as signals, and the simple feature is sent to a convolutional neural network module for identification. The identification method comprises the steps of: modulating a transmittedsignal and performing pulse shaping; performing up-conversion on the transmitted signal and then transmitting the transmitted signal through an additive white Gaussian noise channel; performing pre-processing first by a receiving end to obtain the co-directional component r(t) of the analyzed signal; constructing the simple feature, i.e., constructing the co-directional component r(t) and the quadrature component of the analyzed signal into a two-dimensional matrix; performing feature learning and classification by the convolutional neural network; and sending a modulation method to a demodulation end to obtain a demodulated signal. The method is low in feature design complexity, avoids explicit feature extraction, has high classification correctness, and can be applied to communication systems having high recognition performance requirements.

Description

technical field

[0001] The invention belongs to the field of communication technology, mainly relates to the technical field of communication signal identification and deep learning, specifically a communication signal modulation method identification method based on convolutional neural network, which is used for cognitive radio, interference identification, spectrum management, electronic countermeasures, etc. field. Background technique

[0002] Communication signal modulation identification is an intermediate step in signal detection and demodulation. It has important research value in both military and civilian fields, and is widely used in cognitive radio, interference identification, spectrum management, electronic countermeasures and other fields.

[0003] At present, in the literature on modulation recognition that has been published at home and abroad, the modulation recognition methods can be divided into two categories: decision pattern recognition and statistica...

Examples

Embodiment 1

[0029] Modulation recognition is an important application of pattern recognition in the field of communication. Although many achievements have been achieved by using the traditional method of combining eigenvalues ​​and trainable classifiers, these advances cannot cover up the shortcomings in its development: the existing The extraction steps of the characteristic parameters of the modulation method recognition algorithm are complicated, the features required for system recognition are almost all manually designed, and the recognition rate is low under low signal-to-noise ratio. The present invention conducts research and discussion on the above-mentioned problems, and proposes a communication signal modulation mode recognition system based on convolutional neural network. According to the transmission direction of the signal, it is connected with the system sending end, channel, and receiving end in turn, and the signal is sequentially passed through the above-mentioned units....

Embodiment 2

[0033] The overall composition of the modulation mode recognition system based on the convolutional neural network is the same as in Embodiment 1. The convolutional neural network of the present invention is a multi-layer deep convolutional neural network, including an input layer, four convolutional layers and four down-sampling layers , two fully connected layers and an output layer. see figure 2, in this example, the input layer is a 2×512 matrix. After the first convolutional layer, that is, after 32 1×3 convolution kernels perform convolution operations on the input data, the input data changes to 32 2× 512 feature maps; in order to prevent data overfitting during the training process, the present invention requires 1×2 downsampling operation for 32 2×512 feature maps, and 32 2×256 feature maps are obtained after downsampling operation . In order to fully extract the features in the input data, in this example, the number of convolutional layers and downsampling layers...

Embodiment 3

[0035] The present invention is also a modulation mode recognition method based on a convolutional neural network, which is used on a modulation mode recognition system based on a convolutional neural network. The overall composition of the modulation mode recognition system based on a convolutional neural network is the same as in Embodiment 1-2.

[0036] see figure 1 , the identification method includes the following steps:

[0037] (1) The sending end modulates the sending signal and performs pulse shaping: at the sending end of the modulation mode recognition system based on convolutional neural network, the sending communication signal is modulated according to different constellation maps, and the rectangular pulse is used for pulse shaping operation to obtain a rectangular pulse Shaped signal. The symbol length is selected when performing different modulation modes on the signal, and the symbol length M=32 is selected in this example. The length of the input data in t...

technical field

[0001] The invention belongs to the field of communication technology, mainly relates to the technical field of communication signal identification and deep learning, specifically a communication signal modulation method identification method based on convolutional neural network, which is used for cognitive radio, interference identification, spectrum management, electronic countermeasures, etc. field. Background technique

[0002] Communication signal modulation identification is an intermediate step in signal detection and demodulation. It has important research value in both military and civilian fields, and is widely used in cognitive radio, interference identification, spectrum management, electronic countermeasures and other fields.

[0003] At present, in the literature on modulation recognition that has been published at home and abroad, the modulation recognition methods can be divided into two categories: decision pattern recognition and statistica...

Examples

Embodiment 1

[0029] Modulation recognition is an important application of pattern recognition in the field of communication. Although many achievements have been achieved by using the traditional method of combining eigenvalues ​​and trainable classifiers, these advances cannot cover up the shortcomings in its development: the existing The extraction steps of the characteristic parameters of the modulation method recognition algorithm are complicated, the features required for system recognition are almost all manually designed, and the recognition rate is low under low signal-to-noise ratio. The present invention conducts research and discussion on the above-mentioned problems, and proposes a communication signal modulation mode recognition system based on convolutional neural network. According to the transmission direction of the signal, it is connected with the system sending end, channel, and receiving end in turn, and the signal is sequentially passed through the above-mentioned units....

Embodiment 2

[0033] The overall composition of the modulation mode recognition system based on the convolutional neural network is the same as in Embodiment 1. The convolutional neural network of the present invention is a multi-layer deep convolutional neural network, including an input layer, four convolutional layers and four down-sampling layers , two fully connected layers and an output layer. see figure 2, in this example, the input layer is a 2×512 matrix. After the first convolutional layer, that is, after 32 1×3 convolution kernels perform convolution operations on the input data, the input data changes to 32 2× 512 feature maps; in order to prevent data overfitting during the training process, the present invention requires 1×2 downsampling operation for 32 2×512 feature maps, and 32 2×256 feature maps are obtained after downsampling operation . In order to fully extract the features in the input data, in this example, the number of convolutional layers and downsampling layers...

Embodiment 3

[0035] The present invention is also a modulation mode recognition method based on a convolutional neural network, which is used on a modulation mode recognition system based on a convolutional neural network. The overall composition of the modulation mode recognition system based on a convolutional neural network is the same as in Embodiment 1-2.

[0036] see figure 1 , the identification method includes the following steps:

[0037] (1) The sending end modulates the sending signal and performs pulse shaping: at the sending end of the modulation mode recognition system based on convolutional neural network, the sending communication signal is modulated according to different constellation maps, and the rectangular pulse is used for pulse shaping operation to obtain a rectangular pulse Shaped signal. The symbol length is selected when performing different modulation modes on the signal, and the symbol length M=32 is selected in this example. The length of the input data in t...

Images