Super-resolution Reconstruction Of Spectral Sensing Data From Low Orbit Satellites

The Internet of things has a broad development prospect.The cooperation between the satellite Internet of things and the ground Internet of things can provide better services.When choosing satellite orbit,LEO satellites have the advantages of being light and small,with lower delay and lower cost compared with the high orbit(GEO)satellites.However,in the low orbit satellite Internet of things system,the communication and spectrum sensing common antenna rf equipment,the low orbit satellite can only perceive the spectrum in the communication gap,so the sensing gap increases with the saturation of the communication service.The multi-dimensional spectrum data obtained by using several low-orbit satellites as sensing sources are incomplete and the spatial resolution is low,which results in the low spatial and temporal resolution of the generated spectrum situation and spectrum map.Aiming at this problem,the paper systematically studies the electromagnetic environment map reconstruction technology based on spatial interpolation and the image super-resolution reconstruction algorithm,and respectively completes the spectrum sensing data of low-orbit satellite and improves the spatial resolution of the spectrum sensing data.The work of the thesis is as follows:(1)This paper is based on the electromagnetic perception low-orbit constellation system for the Internet of things.First,the low-orbit Satellite constellation scenario is constructed based on Satellite Tool Kit(STK)to generate spectrum perception of low-orbit satellites.According to the operational characteristics of the satellite and the derived data samples,the analysis results show that there is a large amount of missing sensing data,the longitude and latitude interval between adjacent data is large,and the variation amplitude of longitude is inconsistent with that of latitude.In order to facilitate the processing of subsequent spectrum data,more spectrum sensing data was obtained by combining the operating position of low orbit satellite derived by STK,the antenna azimuth map of earth station and other parameters and the wireless channel characteristics of satellite communication.(2)due to the low-orbit electromagnetic perception constellation system oriented to the Internet of things,low-orbit satellites can conduct spectrum perception in a short time,so the perception data presents the characteristics of sparse spatial distribution and missing quantity,and it is necessary to complete the unknown data through the known data.Combination of comprehensive research on spectrum situation generated the spectrum situational completion algorithm and electromagnetic environment map reconstruction algorithm,based on Leo satellite frequency spectrum perception of time can get the number of data and data characteristics,combined with the common data completion algorithm application scenarios,eclectic improved the classical inverse distance weighted interpolation method and the improved inverse distance weighted interpolation method.(3)a non-uniform spectral coordinate system was established according to the inconsistency of the latitude and longitude of the perceived data,and the spectrum was transformed into a two-dimensional image by associating the spectral coordinate system with the pixel coordinate system,so as to improve the spatial resolution of the spectral perception data by using the image super-resolution reconstruction algorithm.In order to solve the problem of low spatial resolution of low orbit satellite spectrum sensing data,the traditional image super-resolution reconstruction algorithm is investigated based on the amount of spectrum sensing data obtained.In order to improve the spatial resolution of low-orbit satellite spectrum sensing data,the traditional interpolation method,the bayesian method and the self-similarity learning method are adopted respectively.The spatial super-resolution reconstruction of spectrum sensing data in one and two signal sources is analyzed.