Recognition Research For Commonly Used Digital Modulation Signal

With the development of communication technology, space electromagnetic environment becomes more complex, utilizing limited frequency resources reasonablely becomes more and more important, the difficulty of ensuring the correct transmission of information is also bigger, we are faced with great challenge in communication countermeasure field. The communication modulation techniques play important roles both in military and civilian application, which have been widely used in many aspects: signal detection, spectrum supervision, security surveillance, wireless warning, keeping secret,and etc. The modulation recognition is to identify the modulation type of the received signals with little priori information. Modulation recognition technology plays an important role in the field of communication, more attention has been paid, therefore, many scholars have carried on the research to it.In this paper, the modulation recognition are studied based on previous works associating actual projections. The main works and contributions are summarized as follows:Firstly, for digital phase modulation and frequency modulation recognition, the box dimension of the instantaneous amplitude of the modulated signals’ absolute value is used as MPSK and MFSK signals’ characteristic parameter that can identify the modulated classes, to identify the modulation signals in the class, the box dimension of the instantaneous phase of MPSK signals and the box dimension of the instantaneous frequency of MFSK signals are used as the characteristic parameters. To improve the accuracy of recognition, the SNR is estimated, the decision threshold that is a fitting curve related to SNR is proposed, then according to the SNR value, current decisions is determined, finally the recognition of all kinds of modulation signals is realized, this method improves the recognition efficiency under low SNR.Secondly, for the modulation recognition problem of a single modulation signal andseveral time-frequency overlapping modulation signals, the new identification method that combine signals’ cyclic spectrum with sparse representation is proposed. Signals’ cycle spectrum modulation is obtained, and then the cycle spectrum for each class signals are expressed by sparse representation, thus all types of modulation signals’ sparse coefficients are calculated, the sparse coefficients are regarded as identifying characteristics, finally,according to the sparse coefficient and support vector machines(SVM), classification of the modulation signals is completed.