Research Of Algorithms For Modulation Recognition And Parameter Estimation In Non-cooperative Communication

Modulation recognition technology is developed to identify the modulation types by extracting effective characteristic parameters of signals according to little prior information after intercepting the communication signals,which plays a very important role in military and civil fields. However, with the rapid development of communication technology, more and more modulation types have been applied in practical communication systems, and the electromagnetic environment is becoming increasingly complex,the difficulty of modulation recognition is also increasing. Therefore, it is of great significance to study more practical and effective modulation recognition methods. And modulation recognition technology implementation process is inseparable from the support of parameter estimation techniques,the study of parameter estimation techniques is particularly necessary.Firstly, the mathematical models and the modulation parameters of signals which are common in communication systems are summarized,and the theories of communication signals processing methods are introduced briefly; Secondly, wavelet transform, higher order cumulant,higher moments of envelop and transient characteristics analysis are utilized to identify the relevant communication signals,and Mexihat wavelet transform method is proposed to overcome the defects of the traditional Haar wavelet method which is not suitable for the signals after the shaping filter; Thirdly, several classical carrier frequency estimation methods are introduced, and the performance of carrier frequency estimation method based on the cyclic spectrum is verified by the simulation result, the symbol rate is estimated using Mexihat wavelet transform which have been mentioned earlier in this paper, the simulation results verify the accuracy of the estimation algorithm; Finally,this paper designs a real signals acquisition experiment using two Universal Software Radio Peripheral (USRP)devices to validate the effectiveness of the algorithm, and completes the implementation of the algorithm and the design of the interface. the experimental results show that the modulation recognition and parameter estimation scheme in this paper also have a good performance in real systems.