Graph Signal Processing Techniques for Analyzing Aviation Disruptions

Understanding the characteristics of air-traffic delays and disruptions is critical for developing ways to mitigate their significant economic environmental impacts. Conventional delay-performance metrics reflect only magnitude incurred flight at airports; in this work, we show that it also important characterize spatial distribution across a network airports. We analyze graph-supported signals, leveraging techniques from spectral theory graph-signal processing compute analytical simulation-driven bounds identifying outliers distribution. then apply these methods case airport-delay networks demonstrate applicability our by analyzing U.S. airport 2008 through 2017. perform an airline-specific analysis, deriving insights into delay dynamics individual airline subnetworks. Through highlight key differences between different types disruptions, ranging nor’easters hurricanes outages. examine interactions subnetworks system-wide compile inventory outlier days could guide future aviation operations research. In doing so, how approach can provide operational air-transportation setting. Our analysis provides complementary metric conventional aviation-delay benchmarks aids airlines, traffic-flow managers, transportation-system planners quantifying off-nominal system performance.