Research And Realization Of Signal Detection And Recognition Technology Based On Cyclic Stationary Feature

The rapid development of information technology constantly stimulates various types of electromagnetic communication equipment to be updated and iterative,accompanied with increasingly complex and diverse communication signal system and modulation mode,bringing endless emergence in the processing methods.Among a number of communication signal processing methods,cyclostationary signal processing technology stands out due to its pure and efficient processing.The special capability to effectively separate noise and signal in a low signal to noise ratio(SNR)environment is the most significant advantage,based on which this paper studies the signal detection and recognition technology around the characteristics of signal cyclostationarity.Considering the existence of some shortcomings in the traditional cyclostationary feature detection and modulation recognition mechanism,referring to blind cyclostationary signal detection,complex representation of signal cyclic statistics,and incomplete feature extraction of signal cyclostationary in the recognition process,the following related work has been carried out in this paper:(1)Given the weak feature detection performance of the traditional cyclic spectrum without prior knowledge,inspired by the conventional beamformer,the multi-antenna beamforming technology is introduced,forming a blind detection method based on multi-antenna beamforming.This method can be carried out at the receiving end of the antenna through the beamforming technology without prior knowledge(cycle period),with the corresponding test statistics formed.Finally,the signal test statistics are judged by the generalized likelihood ratio test.The simulation results demonstrated its robust signal detection performance in the unknown signal prior information environment and low SNR.(2)Aiming at the existing shortcomings of recognition algorithms based on signal cyclic spectra,such as incomplete feature extraction and poor robustness against noise.A signal recognition method based on a deep residual network and noise reduction cyclic spectrum is discussed in this paper,where the Wiener filtering is used to preprocess the two-dimensional cyclic spectrum of the signal.Then the preprocessed is input into the modified 50-residual layer network for feature extraction and identification of the modulation type of the signal.Experimental results report that the proposed method can effectively identify eight different modulation types of cyclostationary signals,with the recognition accuracy reaching 96%when SNR is greater than-2dB,and the recognition performance is better than some existing deep learning recognition algorithms,such as AlexNet,GoogleNet,etc.(3)In terms of hardware,the Zedboard+AD9361 software radio platform is utilized here to establish a cyclostationary signal recognition system,testing and verifying the signal recognition method based on the deep residual network and the noise reduction cyclic spectrum,with the supporting hardware support package connected.The experimental results indicate that the system can complete the signal recognition task well and has certain theoretical significance and practical value.