Delineation of GNSS Interference Impact Areas Using a Gaussian-Laplacian Multi-Kernel Fusion Method Based on Anomalous Data Features

As a critical infrastructure in modern aviation， the Global Navigation Satellite System （GNSS） is vulnerable to various forms of interference， which potentially cause operational disruptions including flight delays， reduced operational efficiency， and compromised aviation safety. This study addresses current demands by proposing a novel Gaussian-Laplacian multi-kernel fusion method based on Automatic Dependent Surveillance-Broadcast （ADS-B） data to delineate GNSS interference-affected areas. The research methodology involves a comprehensive analysis of abnormal patterns in navigation quality indicators and track information under GNSS interference conditions. Through systematic examination of these data characteristics， we successfully identified distinct anomalous flight data and acquired complete flight trajectory information for corresponding aircraft， thereby constructing a specialized dataset for interference analysis. The proposed method combines Gaussian and Laplacian kernel density estimates via weighted summation， enabling the adaptive capture of both local and broader data distribution characteristics. Using the extracted anomalies， a gridded airspace analysis was performed to quantify abnormal risk levels per grid cell. Interference-affected zones were identified and delineated by comparing these values with a preset threshold. In addition， the effects of the kernel weights and bandwidth on the zoning results were systematically evaluated. Experimental results demonstrated that the multi-kernel fusion approach outperformed single-kernel methods overall. Therefore， this method can assist air traffic management in locating interference-affected areas and initiating timely countermeasures， thereby offering practical technical support for mitigating GNSS interference in aviation.