Abstract: Flight inspection is a necessary technical means to ensure aviation transportation safety. As an important all-weather precision approach and landing guidance system, Microwave Landing System (MLS) lacks comprehensive standards for flight inspection and test evaluation in the civil aviation domain. Therefore, this study investigates MLS flight inspection based on international standards. Addressing the challenges posed by MLS's wide coverage and flexible approach paths, which may result in spatial signal omissions due to limited flight maneuvers covering the paths, we propose a sparse signal-based scene reconstruction method. This approach employs RBFNN to achieve three-dimensional inversion reconstruction of signal scenes under sparse sample conditions, and enables evaluation of the three-dimensional spatial signal scenes of MLS. Additionally, we conduct simulation experiments using the path following error as the sample space. The experimental results show that the reconstruction error of the path following error after three-dimensional inversion reconstruction is about 1.4 feet, which is far less than the 20 feet tolerance required by the FAA8200 flight inspection specification manual. This paper provides a new method for comprehensive evaluation of complete spatial signal scenes in MLS flight inspection.