Research And Implementation Of Spectrum Detection System For LEO Satellite Communication

In pace with evolution of wireless communication technology and expansion of the Internet,satellite communication has ushered in industrial and technological changes.In the near future satellite communication will be integrated with 5G,air communication networks,terrestrial cellular networks,marine communication networks and artificial intelligence.These trends will make satellite communication technology play a more important role in more fields.Especially in recent years,the low-earth-orbit satellite communication system has developed rapidly,and the competition for orbital resources and frequency resources has become increasingly fierce.In terms of key technologies in satellite communication,there is a big gap between domestic level and foreign advanced level.In order to take the lead in the intense global satellite communications competition,the establishment of a low-earth-orbit satellite constellation system for electromagnetic perception is of great significance for managing the spectrum situation and mastering space electromagnetic security.The spectrum monitoring system is an important component of electromagnetic environment perception.At present,the satellite signal spectrum monitoring system mostly relies on human observation,which has low work efficiency and limited functions.The estimation algorithms of signal parameters in the monitoring system also have some defects,such as high complexity,low accuracy,and performance deterioration in the presence of frequency offset.Aiming at the existing problems,this paper designs a satellite signal spectrum monitoring system based on software radio platform,delves into the key technologies in the monitoring software,and finally implements the spectrum monitoring system and tests it.The key algorithms in the satellite communication spectrum monitoring system studied in this paper include the detection of signal,the estimation of the signal carrier frequency,the estimation of the signal symbol rate,the estimation of the signal-to-noise ratio and the auto classification of the signal modulation.Regarding the estimation of the carrier frequency,the centroid detection method,the instantaneous phase method and the extremum-based method are studied in this paper,an improved extremum-based algorithm and an estimation algorithm based on wavelet transform are proposed,which achieve better signal detection probability performance and estimation performance.Regarding symbol rate estimation,in this paper,we study delay multiplication method,wavelet transform method,cyclic spectrum method and vector velocity method,analyze their estimation performance and frequency offset resistance,and propose an improved vector velocity method,which improves the estimation accuracy of vector velocity method under low SNR.In the discussion of auto modulation classification algorithm,a classification algorithm based on statistical features is studied,and three statistical features of the signal are discussed,namely,instantaneous feature,high-order cumulant and quasi-high-power discrete spectrum.Based on them,a decision tree is designed,and the average probability of correct classification can reach more than 97%on average when theE_b/N₀ is above 4d B.In the SNR estimation algorithm,two no-data-aided estimation algorithms are studied,the method based on second-fourth moment and the method based on subspace decomposition.The simulation results show that the subspace decomposition method not only has better estimation performance but also is suitable for constant envelope and non-constant envelope modulation signals.In the specific implementation of satellite signal spectrum monitoring system,the software radio platform USRP X310 is used to monitor and collect satellite signals,and the Lab VIEW development environment is used to develop software client to meet different functional requirements of the monitoring system.