Shortest Path Algorithms: An Evaluation Using Real Road Networks

Abstract

The classic problem of finding the shortest path over a network has been the target of many research efforts over the years. These research efforts have resulted in a number of different algorithms and a considerable amount of empirical findings with respect to performance. Unfortunately, prior research does not provide a clear direction for choosing an algorithm when one faces the problem of computing shortest paths on real road networks. Most of the computational testing on shortest path algorithms has been based on randomly generated networks, which may not have the characteristics of real road networks. In this paper, we provide an objective evaluation of 15 shortest path algorithms using a variety of real road networks. Based on the evaluation, a set of recommended algorithms for computing shortest paths on real road networks is identified. This evaluation should be particularly useful to researchers and practitioners in operations research, management science, transportation, and Geographic Information Systems.

Average-case complexity of single-source shortest-paths algorithms: lower and upper bounds

“…Hence, at first glance, label-setting algorithms seem to be the better choice. However, several independent case studies [6,10,13,28,29,44,50,70] showed that SSSP implementations of label-correcting approaches frequently outperform labelsetting algorithms. Thus, worst-case analysis sometimes fails to bring out the advantages of algorithms that perform well in practice.…”

“…However, none of the alternative label-correcting algorithms succeeded to asymptotically improve on the O(n · m)-time worst-case bound of the simple BellmanFord approach. Still, a number of experimental studies [6,10,13,28,29,44,50,70] showed that some recent label-correcting approaches run considerably faster than the original Bellman-Ford algorithm and even outperform label-setting algorithms on certain data sets. So far, no profound average-case analysis has been given to explain the observed effects.…”

“…Initially, Q 1 is empty and Q 2 holds the source node s. If tent(v) is reduced due a relaxation of the edge (u, v) while scanning u then v is only added to either Q 1 or Q 2 if it is currently not stored in any of them: if v was scanned before then it will be added to the tail of Q 1 , or to the tail of Q 2 otherwise. This approach has worst-case execution time O(n 2 · m) but performs very well on many practical inputs [6,70].…”

Energy-aware management for cluster-based sensor networks

“…Our routing problem can be considered as the transpose of the one-to-some shortest path, since not all sensors are active simultaneously. A recent study by Zhan and Noon [33] suggested that the best approach for solving the one-to-some shortest path is Dijkstra's algorithm. In addition, Dijkstra's algorithm is shown to suit centralized routing [28].…”

A Novel Route Guidance Algorithm with Maximum Coverage and Minimum Handover for Vehicular Networks

“…In [3], authors evaluate 15 shortest path algorithms in real road networks of 10 states of the United Sates. They identify a set of recommendations regarding algorithm selection based on the computational performance.…”