Modulation Recognition And Parameter Estimation Based On Cyclic Spectrum

Modulation recognition has been widely used in the military and civilian fields.In the military,modulation recognition is the basis for electronic warfare measures and methods such as electromagnetic interference implementation,electromagnetic threat assessment and analysis,and electromagnetic environment monitoring;in the civilian field,such as software-defined radio and cognitive radio system,modulation recognition also plays a very important role in these intelligent communication systems.Cyclic spectrum has many advantages as a signal analysis tool.It requires low prior knowledge,provides a richer analysis domain,can effectively reduce the influence of stationary noise and interference signals,and contains more parameter information than traditional spectrum analysis methods.Take electromagnetic environment of the battlefield as the research background the thesis focuses on the modulation recognition and parameter estimation of common digital signals.The main research content of the thesis includes the following three parts:(1)Detection and type identification of pulse radar signals and communication signals.The receiver in a complex electromagnetic environment may receive either pulse radar signals or communication signals.For this reason,frequency domain characteristics are extracted to detect whether there exists radar or communication signals.Then the difference in the duty cycle of the two types of signals are extracted to identify the type of the received signal.The simulation results show that the target signal can be detected when the signal-to-noise ratio is small,and the signal type identification can achieve high accuracy.(2)Recognition of the modulation types of three digital communication signals.The cyclic spectrum characteristics of three commonly used digital communication signals of BPSK,QPSK and MSK are analyzed in detail.Then several feature parameters extracted from cyclic spectrum which are commonly used in related literature,are analyzed and their advantages and disadvantages are discussed.Then two feature parameters based on ratio of magnitude are derived.These parameters are easy to extract and has low computational complexity.Finally,appropriate thresholds are designed and an easy-to-implement decision tree classifier is used to identify the modulation types of the above three signals.The modulation recognition scheme designed in this thesis does not require too much preprocessing such as synchronization and prior information such as carrier frequency and symbol rate.The features used are easy to extract and have good robustness.Simulation show that recognition results of modulation type are accurate.(3)Parameter estimation of three types of digital communication signals.Cyclic spectrum contains parameter information that can be exploited to estimate the symbol rate and carrier frequency of BPSK,QPSK and MSK.The normalized mean square error is introduced to evaluate the performance of parameter estimation,and the estimation performance under different data lengths is compared.Simulation results show that high estimation accuracy can be achieved even when the signal-to-noise ratio is small.