Spatial Data Map Generation for Wireless Networks

The patent divides the wireless network coverage area into discrete spatial units called 'tiles' or grid cells. Each tile independently stores and processes spatially dependent information such as KPIs, trust scores, and derived metrics. This segmentation enables systematic organization of spatial data while maintaining adaptability to local transmission conditions and impairments in each tile region.

The patent introduces an intermediary processing layer that receives raw network data from multiple access points, computes derived information through iterative evaluation, and organizes it into structured spatial maps. This intermediary layer bridges the gap between raw data collection and organized spatial information presentation, enabling both adaptability to impairments and structured organization.

The patent merges data from multiple access points and client devices into a unified spatial map structure. By combining measurements and KPIs from numerous data sources into a single integrated spatial representation, the system achieves comprehensive network insights while managing complexity through unified data organization rather than separate processing for each source.

The system employs self-organizing algorithms that automatically process and structure incoming network data without requiring complex manual configuration. The spatial map structure naturally organizes itself as data is collected, with tiles automatically updating their content based on received measurements, reducing the operational complexity of managing multiple data sources.

The patent pre-structures the spatial domain into a grid of tiles before data collection begins. This preliminary organization allows interpolation operations to be confined to local tile boundaries rather than processing entire network areas, significantly reducing computational time while maintaining continuous spatial information through localized kriging operations within each tile.

The system applies kriging interpolation selectively only to tiles that contain missing data points, rather than processing all tiles uniformly. This partial action approach maintains continuous spatial information where needed while avoiding unnecessary computational overhead in tiles that already have complete data, reducing overall processing time.