Research On The Modulation Recognition Algorithm For Single-channel Time-frequency Aliasing Signals

With the development of modern communication technology,the electromagnetic spectrum resources are becoming scarce gradually,and the complex electromagnetic environment which is dense in time domain and aliasing in frequency domain is formed.When the single-channel receiving method is adopted,it is inevitable to receive time-frequency aliasing signals.And the modulation recognition algorithm of single-channel time-frequency aliasing signals need to be studied.Most of algorithms are only applicable to the case where the aliasing signals are of different carrier frequency,but not suitable for signals of the same carrier frequency.Therefore,in this paper,a modulation recognition algorithm for single-channel time-frequency aliasing signals based on the joint feature extraction of high-order cumulants and instantaneous features is proposed,to make up for deficiencies of previous algorithms.When the carrier frequency is the same,those algorithms are studied on only few types of signals,and they are difficult to be recognized,affected by noise greatly,have poor timeliness,and need more prior knowledge and so on.With the same carrier frequency of the aliasing signals,the algorithm in this paper complete the recognition and classification of 21 kinds of time-frequency aliasing dual signals which are composed of any two combinations of {4ASK,BPSK,QPSK,8PSK,16 QAM,2FSK}.Firstly,the basic principles of several digitally modulated signals is introduced in this article.Then,the mathematical model of the time-frequency aliasing signal received by single channel is given.In addition,to solve the problem of poor universality of previous definition of aliasing degree,a new definition method for broader aliasing degree is proposed.After that,the basic theory of modulation recognition feature extraction is introduced,including the principles and properties of higher-order cumulant features and instantaneous features,which lays theoretical foundation for future research.Secondly,this paper studies the feature extraction algorithm of time-frequency aliasing dual signals composed of any two of the six common digital modulation signals {4ASK,BPSK,QPSK,8PSK,16 QAM,2FSK}.High-order cumulant feature extraction algorithm has been proposed firstly,and 4 kinds of features were extracted,which can complete the main classification of aliasing signals.And then the feature extraction algorithm based on instantaneous features was studied,which extracted 3 kinds of instantaneous features to complete the classification of remaining signals.After theoretical analysis and simulation verification,the high-order cumulant features and instantaneous features extracted in this paper can complete the recognition and classification of 21 kinds of aliasing signals.Finally,the recognition algorithm based on joint features is proposed in this paper,and the decision tree classifier and support vector machine classifier are constructed,then the process design and performance analysis on the recognition algorithm are performed.The algorithm using support vector machine classifier has better recognition effect under low signal-to-noise ratio.In addition,the influence of aliasing ratio and power ratio on the recognition performance of the algorithm is studied.Simulation results show that the algorithm in this paper has a certain robustness to the aliasing degree,and the algorithm performance is not affected when the power ratio is lower than 3:1.The algorithm proposed in this paper is robust.