Faster algorithms for the shortest path problem

Abstract

We investigate efficient implementations of Dijkstra's shortest path algorithm. We propose a new data structure, called the redistributive heap, for use in this algorithm. On a network with n vertices, m edges, and nonnegative integer arc costs bounded by C, a one-level form of redistributive heap gives a time bound for Dijkstra's algorithm of O(m + nlogC). A two-level form of redistributive heap gives a bound of O(m + nlogC/loglogC). A combination of a redistributive heap and a previously known data structure called a Fibonacci heap gives a bound of O(m+ nsqrt{log C}). The best previously known bounds are O(m + nlogn) using Fibonacci heaps alone and O(mloglogC) using the priority queue structure of Van Emde Boas, Kaas, and Zijlstra.