Research On Modulation Recognition Method Of Signals For Underlay Spectrum Sharing

In the cognitive radio technology of underlay,the secondary users can communicate with the same spectrum while ensuring the communication of the primary users safety and reliability,and the spectrum utilization is greatly improved compared to the case of allocating spectrum only to the primary users.It has become one of the important research contents to solve the current shortage of spectrum resources.However,in the interference temperature range,underlay mode is that multiple users send message in the same channel at the same time.So,signals are time-frequency overlapped in the channel,and the received signal model is more complicated.In signal processing research,modulation recognition is the process of judging the modulation type of the received signals,indirectly affecting the value of the interference temperature,provides a reference for subsequent signal demodulation,and is of great significance for research on spectrum management,illegal signal monitoring,etc.This paper is concerned with the modulation recognition for signals in underlay cognitive radio.The specific research contents are as follows:1.In order to solve the low recognition rate of the time-frequency overlapped MQAM modulation method in underlay cognitive radio,a novel method of time-frequency overlapped signals modulation identification based on time-frequency analysis of image and gray-gradient co-occurrence matrix is proposed.First,after the frequency slice wavelet transform(FSWT)for the received time-frequency overlapped MQAM,the time-frequency analysis images are obtained.Then the slices of the image with obvious texture difference are selected for grayscale processing,and the gray-gradient co-occurrence matrix of the time-frequency analysis image is calculated as the feature vector.Finally,the probabilistic neural network classifier is used to effectively realize the modulation recognition of the time-frequency overlapped MQAM signals.For verifying the effectiveness of the proposed method,experimental simulations are carried out in terms of different signal-to-noise ratios(SNR)and number of signals,different spectrum overlapped rates,and different power ratios.The results show that when SNR is 4 d B the average recognition rate of the scheme can reaches 95% or more;the proposed method is robust to both the power ratios and spectrum overlapped rates of the component signals.In addition,in the comparison experiment,when SNR is greater than 0 d B,the modulation identification performance of the proposed method is superior to the traditional methods.2.In underlay cognitive radio,this paper proposes a method for modulation recognition of time-frequency overlapped MPSK/MQAM signals based on the contour maps of cyclic spectrum and convolutional neural network(CNN)models.Aiming to achieve modulation recognition of the signals,the proposed method extracts the initial features of the received signals,and then the CNN model is used to extract further features as the final features.The specific steps are: First,calculate the contour maps of cyclic spectrum of the received signals;after preprocessing the contour maps to serve as input data for the CNN model,construct and optimize the CNN model;then complete the training of the CNN models;finally,the well trained CNN model is used for modulation recognition of the signals.Simulation results show that when the SNR is 2d B,the average recognition rate of the proposed method is 90%,and it is robust to the power ratio of the component signals and spectrum overlapped rates of the component signals.In the comparison experiment,although the proposed method is less effective than the first method,the modulation recognition performance of this scheme is better than the traditional methods when SNR is greater than 0 dB.