Modulation Identification Method Based on Differential Density Constellation

Abstract

The invention discloses a modulation identification method based on a differential density constellation diagram, which determines the modulation mode to be identified from PSK modulation and QAM modulation according to actual needs, collects signal samples of these modulation modes, and generates a corresponding differential constellation diagram; Set the PSK color sequence and QAM color sequence; firstly, gather points for the differential constellation diagram, and then divide it into several grids, count the data point density in each grid, and generate the difference according to the predetermined range and color sequence of the data point density Density constellation diagram; the neural network is trained by using the differential density constellation diagram of the signal sample and the modulation mode label; when performing modulation identification, the differential constellation diagram of the modulation signal to be identified is first clustered, and the modulation is initially judged according to the number of cluster centers way, and then generate a differential density constellation, which is input into the neural network to obtain the modulation recognition result. The invention adopts the differential density constellation diagram combined with the neural network to effectively improve the modulation recognition rate.

Description

technical field [0001] The invention belongs to the technical field of modulation identification, and more specifically relates to a modulation identification method based on a differential density constellation diagram. Background technique [0002] Modulation recognition is an important technology between signal detection and signal demodulation, and is one of the key technologies in software radio and non-cooperative communication. The prerequisite for successful analysis and demodulation of communication signals is to know the modulation style and characteristic parameters of the signal. The modulation style is an important basis for distinguishing different modulated signals, and the blind recognition technology of modulated signals can pass through the receiver without prior knowledge. From signal processing to judging the type of received signal, it provides a huge help for the next step to analyze the characteristics of the signal [0003] At present, according to t...

AI Technical Summary

Problems solved by technology

However, most of the network models used in modulation recognition algorithms based on deep learning are traditional CNN network models, such as AlexNet, VGG16, etc. These networks are mostly designed and constructed with image recognition as the main background. The network model is large, the number of parameters is large, and time-consuming It takes a long time, the network performance is wasted, and it is not suitable for the modulation recognition field

At the same time, most of the current network identification data are IQ signals and constellation diagrams. Data characteristics are not fully considered, and the noise resistance is poor. At the same time, network performance is not fully utilized, and the identification effect is not ideal.

Images