The Research On Technologies Of Compression Storage For Trajectory On Road Network

The increasing of location aware mobile devices such as vehicle navigationequipment and smart phones has enabled Location-based service (LBS) a reality.Enormous trajectories have been collected by positioning services. Movementtrajectory compression has become an urgent necessity to store these data. However,the current technology can’t ensure the quality of compressed trajectory under highcompression.This paper has proposed a model for trajectory compression based on semanticsin road network. The model can ensure the quality of compressed trajectory underhigh compression ratio. In consideration of the repeatability for trajectories in acertain geographic space, a compression technique has been proposed based on thehistory trajectories. This technique has significant advantage in space savings.Firstly, this paper analyzes the issue of traditional methods for trajectorycompression. To measure the performance of these methods, we proposed the criteriato quantify the compression ratio and the compression error. These methods often getan intolerant error under high compression ratio, and easy affected by the quality ofsampling trajectories.Secondly, this paper has proposed the method named EnHanced SemanticTrajectory Compression, which compresses trajectories based on road semantics aswell as motion features during a road. The method firstly matches each point inoriginal trajectory to a road based on the digital map. Then, the maximum sequence ofpoints belongs to a same road should be chunked to a same road segment. Duringchunking on every road segment, each point in the sequence will be detected if themotion feature changed distinctly. The motion feature change is measured by thevelocity changes. Those points with great feature changes will be recorded in thefeature point list of underlying road segment, and eventually stored in the compressedtrajectory. Points with non-obviously velocity changes will be abandoned. Ourmethod gets an acceptable information loss with a little sacrifice in storage spacecompared to the simple semantic trajectory compression, which represents trajectorieswith only road segments.Thirdly, observing the repeatability for trajectories in space information, thispaper proposed a new technique using history trajectory representing new coming trajectories. The routes within a certain geographical area are limited, as well as thespace information of sampled trajectories. As the amount of trajectories grows, thespace information is easily repeatable. This paper states the repeatability intrajectories, and shows our method’s advantage in compression ratio.In summary, the proposed algorithm EnHanced Semantic TrajectoryCompression has alleviated the conflict between compression ratio and compressionerror. The algorithm ensures the availability of compression trajectories under highcompression ratio. Using history trajectories to represent the new coming trajectoriesmakes it possible to get a very high compression ratio, which makes sense totrajectory compression as well as trajectory mining.