Core Overbounding and its Implications for LAAS Integrity

Empirical studies have observed non-Gaussian behavior in pseudorange correction errors for the Local Area Augmentation System (LAAS). This paper introduces an overbounding technique, called core overbounding, designed to validate integrity for error distributions with heavier than Gaussian tails. The core overbounding process generates a conservative representation of the actual error distribution by decomposing the bound into two parts: an explicit function describing the distribution core and an implicit function describing the far tails. For LAAS applications, it is convenient to express the explicit core bound as a Gaussian Core (GC) or as a Gaussian Core with Gaussian Sidelobes (GCGS). These forms balance bound sharpness, important for heavy-tail mitigation, with operational simplicity, necessary for bound broadcast and protection level establishment in a real-time implementation. Analysis using the GC and GCGS bounds indicates that heavy tails have a relatively weak impact on availability for Category I LAAS but a significant impact for Category II