WorksPhysics and Astronomy
Where are the shortest-path nodes in spatial networks?
Abstract
To predict the spatial distribution of shortest-path nodes, we propose a Gaussian estimate of the shortest-path node probability , the probability that a node lies on a shortest path between a source and a destination . According to the Gaussian estimate, three variables affect the shortest-path node probability and combine into one variable, the standardized detour . In our simulations of hard and soft random geometric graphs in Euclidean, hyperbolic and spherical space, the standardized detour alone approximately determines how fast the shortest-path node probability decreases as the node moves away from the geodesic between the source and the destination . We find and explain that, when the hopcount variance is very small, the shortest-path nodes have a discrete spatial distribution, a quantized distribution of shortest-path nodes. We further extend the study to spatially inhomogeneous networks, for which we propose a method to find the shortest-path nodes.