Computing Shortest Paths Using A*, Landmarks, and Polygon Inequalities (Abstract)

TL;DR

This paper presents a novel heuristic for the A* algorithm that uses polygon inequalities and two landmarks to reduce preprocessing data and accelerate shortest path searches.

Contribution

It introduces a new heuristic based on generalized polygon inequalities that requires less preprocessing data than existing landmark-based methods.

Findings

Faster shortest path queries compared to previous landmark heuristics

Reduced preprocessing storage requirements

Effective use of two landmarks for lower bound computation

Abstract

We introduce a new heuristic for the A* algorithm that references a data structure much smaller than the one required by the ALT heuristic. This heuristic's benefits are permitted by a new approach for computing lower bounds using generalized polygon inequalities, leveraging distance information from two landmarks as opposed to the common single landmark paradigm. In this paper, we demonstrate that this heuristic stores a reduced amount of preprocessing information in comparison to previous landmark algorithms while performing faster search queries.