Efficient route search on hierarchical dynamic road networks

The widespread use of GPS navigations and trip planning on web has aroused considerable interests in fast and scalable path query processing. Existing research has mostly focused on static route optimization where the traffic network is assumed to be stable. Nevertheless, in most cases, route planning is in the presence of frequent updates to the traffic graph due to the dynamic nature of traffic network, and such updates always greatly affect the performance of route planning. Most existing methods, however, cannot efficiently support traffic aware route planning. In this paper, two efficient strategies are proposed to handle this problem. We analyze the traffic condition on the road network and explore spatio-temporal knowledge to guide effective route planning. In particular, several effective techniques are employed to avoid both unnecessary calculations on huge graph and excessive re-calculations caused by traffic condition updates. A comprehensive experiment is also conducted to evaluate the performance of our proposed strategies.