Identification, Based On The Modulation Of Independent Software Radio Receiver

Modulation classification is a very important subject in the non-cooperative communication field, and it has been applied to many fields such as signal identification, interference identification, radio interception and monitoring, etc. the purpose of modulation classification is to decide the modulation type and estimate the signal parameters without any priori knowledge, such as the carrier frequency, symbol rate, etc., and then determine the signal modulation type.The National Aeronautices Space Administration has embarked on an ambitious project to develop new technology named autonomous software-defined radio receiver.The software-defined radio need to reprogramming to defined the functions of the radio when the system parameters are changed. The fundamental difference between a conventional radio is that an autonomous radio has the ability to recognize features of an incoming signal and to respond intelligently, without explicit pre-configuration or reprogramming to define the functions of the radio. The purpose of autonomous radio is to solve problems in deep-space communication. For example, Mars Exploration Rovers is very expensive, so it is used in many years. It is unlikely that they will all use the same data rate, protocols, and modulation types. So it is desired to process the new technology and parameters without pre-configuration. Main content is as follows:Study on the algorithm of the draw the instantaneous characteristic parameters, statistical parameters methods and the power spectrum characteristic. Then simulate these characteristics and analyze results. Design a modulation classification scheme for digital signals, the algorithms is simulated and performance is analyzed.Study on the wavelet transform theory, and implement in the modulation classification for PSK, FSK and 16QAM. Haar mother wavelet is used to draw the wavelet transform amplitude as the classification feature. The simulation results show that it can reach high classification accuracy in low SNR.Study on the concept, structure of the autonomous software-defined radio receiver. It is different from conventional radio. It can receive a signal without much a prior knowledge about its defining characteristics. Some important signal processing modules has been researched such as SNR estimation, rate estimation, symbol timing error estimation and modulation classification. Rate estimation and symbol timing error estimation are based on the SSME algorithm in the SNR estimation module. ML algorithm is used in the modulation classification. And these algorithms are simulated, then analyze results and performance.On the base of the concept and algorithm of the autonomous software-defined radio receiver, a comprehensive modulation classification scheme with parameter estimation is designed. It can work with unknown rate and symbol timing error. Because much of the algorithms are based on the ML theory, a rate pre-estimation module is added for reduce computational complexity. Simulation results show that parameter estimation module has good performance when SNR≥5dB, and modulation classification can reach the 100% accuracy when SNR≥3dB.