A Research Of Trajectory Big Data Compression Technology

The ubiquitousness of GPS sensors in smart-phones,vehicles and wearable devices has enabled the collection of massive volumes of trajectory data from tracing moving objects.However,the scale of trajectory data that grows exponentially has caused a series of problems in the fields of data storage,communication bandwidth,data processing and analysis.Transferring and storing raw trajectory data consumes too much network bandwidth and storage capacity,increases the number of communication between the sensor side and the server side and power consumption.In addition,the query,analysis,and mining of massive raw trajectory data may also cause delays and so on.Consequently,an unprecedented scale of timestamped GPS data has been generated and posed an urgent demand for an effective storage mechanism for trajectory databases.The mainstream compression technique is called trajectory simplification,that finds a subsequence to approximate the original trajectory and attempts to minimize the information loss under a distance measure.Even though various simplification algorithms have been proposed in the past decades,there still lacks a thorough comparison to cover all the state-of-the-art algorithms and evaluate their quality using datasets in diversified motion patterns.Hence,it still remains a challenge for GPS data collectors to determine a proper algorithm in a concrete application.In addition,almost the entire line of previous methods uses error-based metrics to evaluate the compression quality,while ignoring their usability in supporting spatio-temporal queries on top of the reduced database.All in all,the existing research on the field of trajectory compression lacks comprehensive comparative research,experimental analysis,and ignores the compressed data usability.To bridge these gaps,we conduct so far the most comprehensive evaluation on trajectory simplification techniques.According to the experimental findings,we present useful guidance for the selection or development of effective trajectory simplification algorithms.Therefore,in this paper,we conduct so far the most comprehensive evaluation on trajectory simplification,covering 25 algorithms in total and 5 real datasets in different motion patterns.The evaluation covers four types of error metrics.Equally important,we propose to use the data usability of reduced trajectory database as an alternative performance indicator for compression quality and conduct the first experimental study on the accuracies of supporting range queries,6)NN queries,spatial joins and trajectory clustering on top of simplified trajectory database.The experimental results show the following points: 1)With the improvement of hardware performance,offline application scenarios,considering various performances of various algorithms,MRPA proposed by Chen et al,is relatively suitable;the online application scenario,the performance of DOTS proposed by Cao et al is ideal.2)Proposing new error metric can improve the existing trajectory compression technology,providing improved direction for researching more effective trajectory simplification algorithm.For example,MPRA,DOTS performance relatively good,the reason is that a good Integral Square Synchronous Euclidean Distance(ISSD)error metric is applied.3)Although the trajectory simplification algorithm based on the Direction-Aware Distance(DAD)error metric can well preserve the directional information of the trajectory,it cannot capture the position and time information well.Therefore,unless the direction information is key information,such trajectory compression algorithm is not recommended.4)The existing trajectory compression algorithms had poor performance of compression ratio,compression time and compression error.The exits distributed trajectory compression algorithms are not enough to cope with the scale of trajectory data under smart city,so it's urgent to propose new distributed trajectory simplification system using Spark.