Research On Parameter Estimation Method And Application Of Non-cooperative Satellite Based On GPU

After decades of rapid development,the research on communication signal parameter estimation has penetrated into various fields.In the modern combat environment with complex and changeable electromagnetic environment,signal processing is facing unprecedented challenges and pressures,which makes the cooperation of satellites with Africa difficult.The parameter estimation of the signal has extremely important strategic significance.At present,there are few researches on fast parallel estimation algorithm of non-cooperative satellite signal parameters.This thesis mainly studies the estimation method of non-cooperative satellite signal parameters,and uses GPU parallel algorithm to accelerate the signal parameter estimation method in parallel,reducing the signal parameter estimation time,thereby improving the efficiency and accuracy of signal parameter estimation.The main work is as follows:The spectrum detection algorithm of non-cooperative satellite signals is studied.A GPUbased fast frequency-domain local detection algorithm is designed to achieve rapid detection of complex noise floor spectral sub-band signals;a multi-thread-based signal spectral feature extraction algorithm and a GPU-based spectral refinement spectral line detection optimization algorithm are designed.Quickly extract the spectral line features of all subband signals and identify the modulation mode of the detected sub-band signals.Spectrum detection is carried out on the signal data of 250 MHz,150MHz,100 MHz and 50 MHz sampling rate respectively,and the performance and accuracy are analyzed,and more accurate signal detection and identification results are obtained,which effectively improves the complex noise floor spectrum subband signal detection and identification efficiency.The basic theory of continuous estimation of synchronization parameters of non-cooperative satellite signals is studied.A fast tracking algorithm for continuous estimation of synchronization parameters based on GPU parallel is designed.The actual synchronization parameters are continuously estimated and fast tracked for the signal acquisition data with a center frequency of 1575 MHz,a sampling rate of 100 MHz,a center frequency of 1227 MHz,and a sampling rate of 100 MHz,respectively,and the ideal carrier phase tracking deviation and code frequency tracking deviation results are obtained.The signal acquisition data with signal center frequencies of 50 MHz,100MHz,and 250 MHz are selected for performance testing.The results show that the tracking performance of the algorithm is effectively improved.Finally,based on the GPU parallel algorithm,the rapid and continuous estimation process of real-time synchronization parameters is designed and verified with the actual collected data.Combined with the research on the fast identification and parameter estimation of noncooperative satellite signals in this paper,a non-cooperative satellite signal parameter analysis software based on GPU parallel computing is designed and implemented.The software has network communication,remote control of acquisition equipment parameters,The functions of signal modulation mode identification,parameter estimation,spectrum detection,text image curve display of analysis results,and result storage,etc.,realize the effect of fast and accurate analysis of non-cooperative satellite signal parameters.