Research On Blind Detection Technologies Of Low Frequency Communication Signals

With the development of wireless communication technologies,the electromagnetic environment of space becomes more and more complex,and the modulation types becomes the primary characteristic of communication signals.Digital communication signals modulation mode identification,in the military electronic confrontation and civil electromagnetic spectrum monitoring and other fields have played a very important role.The communication signals detection needs to process and judge the identification signals in the complicated electromagnetic environment,and obtains the parameters such as modulation mode,carrier frequency and signals bandwidth of the non-cooperative communication so as to obtain the information,search the signals source and interfere with the communication.Based on the previous research,the detection theory of 2ASK,2PSK,QPSK,2FSK and MSK signals in non-cooperative communication is studied.The detection methods of 2ASK,2PSK,2FSK and MSK are studied.A complete set of system which can detect the existence of the signals,the modulation mode identification and the parameter estimation is designed.The main contents are summarized as follows:Firstly,the cyclic spectral density function is deduced and the cyclic spectrum density functions of the commonly used communication signals are calculated,including 2ASK,2PSK,QPSK,2FSK and MSK.The cyclostationary characteristics of the communication signals and the acyclic of Gaussian white noise are analyzed.And the existence of the discriminant signals is detected in each non-zero cyclic frequency cross section.Secondly,the time-domain characteristics and spectral correlation characteristics of 2ASK,2PSK,QPSK,2FSK,MSK signals are analyzed,and modulation recognition of these five modulation signals is completed.Aiming at the deficiency of single-peak-2FSK signals and MSK signals recognition in the past algorithms,a characteristic parameter based on peak to average ratio of differential phase is proposed to identify single-peak-2FSK signals and MSK signals.The simulation results show that the average correct recognition probability of these two signals is more than 99% when the signals to noise ratio is more than 10 dB,and the whole recognition probability is more than 92%.Thirdly,the relationship between the carrier frequency and the signals bandwidth of the modulation signals based on the zero-frequency normalized cross-section of the cyclic spectral density function is studied,and the carrier frequency estimation and the bandwidth estimation of the modulated signals are carried out by the cyclic spectrum cross-section.The simulation results show that the relative error of carrier frequency estimation is less than 0.500% and the relative error of signals bandwidth estimation is less than 3.525% when the SNR is more than 10 dB.Finally,according to the design of the algorithm,the hardware platform based on the structure of FPGA+DSP solution to implement the algorithm,three kinds of signal hardware system designed by the use of the signal generator for real-time detection,through the serial interface feedback data show that the hardware platform to achieve the desired requirements.