Continuous Predictive Line Queries under Road-Network Constraints

Abstract

With massively available global positioning systems, one can be up to date with its own position even when they are mobile. These position information, collectively, allows serving more knowledge to people on their neighborhood. This paper presents continuously monitoring predictive line query, which provides predicted future traffic information. The information would encourage the user to align his/her journey better, depending on the predicted traffic condition. Naturally, the accuracy of a prediction may become invalidated over time. The proposed query algorithm, thus, considers the continuous monitoring line query which keeps the issuer up-to-date over the time. If there is any significant change in the prediction results on the querying road due to location updates of other vehicles, the updated query result will be automatically sent back to the user. To speed up query processing, a novel data structure is designed, the TPR^Q-tree. The TPR^Q-tree facilitates one update message to be considered on a group of queries. This group wise consideration, contrary to the individual consideration, has reduces the execution time significantly. The results of extensive experimental study demonstrate the efficiency and effectiveness of proposed approach.