An improved shortest route algorithm in Vehicle Navigation System

As a basic function of Vehicle Navigation System, the shortest route algorithm has been a research hotspot in vehicle navigation field. Because of real time requirement of the practical system, it is necessary to optimize Dijkstra algorithm - a typical single-source shortest route algorithm. Based on the analysis of the temporal complicacy and spatial complicacy of Dijkstra algorithm, this paper presents a novel improved Dijkstra algorithm, which can run much more efficiently compared with original algorithm. Firstly, the paper adopts the adjacency list as store structure of topology of road network, and reduces the memory requirements of the algorithm and increases the search speed of the adjacent node. Secondly, binary heap is used to implement the operation of priority queue, which successfully improves efficiency of searching the minimum cost node. Thirdly, search course is divided several phrases according to the special spatial distribution of nodes which have been generated but not extended. A search mechanism of dynamic restricting the search area in stages is introduced, which can gradually reduce search area and greatly decrease the search data volume of the algorithm. Finally, the tests in real road network prove that the modified algorithm is workable and time-saving.