Research On Spectrum Map Reconstruction Algorithm Under Incomplete Observation Conditions

With the continuous development of the information technology industry,the use of spectrum maps to characterize the spatial distribution of signal strength in a particular area has become necessary in spectrum management applications such as frequency reuse and coverage prediction.The spectrum map is usually expressed as the spatial distribution of radio parameters such as received power within the region of interest.Compared with the traditional spectrum sensing technology,the spectrum map can more fully describe the distribution of spectrum resources and facilitate the spatial multiplexing of spectrum resources,and have broad civil and military prospects.How to construct spectrum maps using electromagnetic spectrum monitoring data has important academic and applied research value.Considering the rapid development of national and military electromagnetic spectrum monitoring equipment and stations in recent years,the generation of terabytes of electromagnetic spectrum monitoring data and equipment deployment restrictions make the monitoring data incomplete in the airspace and cannot cover all the locations that need to be monitored.The spectrum map reconstruction algorithm under incomplete observation conditions is studied to explore the spectrum map reconstruction algorithm which can solve the actual demand,and to improve the existing algorithms.The main work of this paper is as follows.First,this paper first reviews the existing methods,and introduces the classic algorithms that can be used to implement spectrum map reconstruction,including matrix completion algorithm and spatial interpolation algorithm.The spatial interpolation algorithm introduces the Kriging interpolation algorithm and the surface spline interpolation algorithm.In the introduction of the Kriging algorithm,the ordinary Kriging and the simple Kriging are distinguished.Secondly,in view of the problem that the spectrum map complement performance is not ideal when there are many radiation sources,a spectrum map based on the difference of observation values is proposed by using the relationship between the number of radiation sources,the rank of the spectrum map and the complement performance.Completion method.The method can be used to realize the iterative completion mechanism of the spectrum map based on the difference of the measured values according to the change of the spectrum map of the adjacent time before and after the change of the number of reconstructed radiation sources,and the reconstruction accuracy and speed can be ensured when there are many radiation sources.The simulation part verifies the accuracy of the proposed algorithm by considering only the path loss and the scene loss considering the path loss and the large-scale fading.The performance of the algorithm and the matrix completion algorithm are compared,and the conclusion is that the algorithm is more accurate.Thirdly,in view of the low accuracy of the ordinary Kriging interpolation algorithm,the Egli radio propagation model information is used to improve the variogram,and the Kriging interpolation algorithm based on the propagation model is obtained.It is proposed to combine the improved Kriging interpolation algorithm with the matrix completion algorithm to improve the defects of the two algorithms.The simulation experiments verify the accuracy of the proposed algorithm by considering only the path loss and the scene loss and large-scale fading scenarios.The performance of the algorithm and matrix complement algorithm and surface interpolation algorithm are compared.The conclusion that the reconstruction performance of the proposed algorithm is higher is obtained.The effectiveness of the proposed algorithm is verified by actual measurements.