HESAVE: An Approach for Online Heuristic GPS Trajectory Sampling

In location based service system, online trajectory compression can help to relieve the whole system’s pressure by reducing communication, storage and computation cost during network transmission, storage and business computing. Segment simplified sampling method is a kind of trajectory compression methods which is widely used in online trajectory compression, but current methods have the problems such as temporal information loss and difficulty of parameter selection. In this paper, we propose an online heuristic trajectory sampling algorithm base on segment simplification, HESAVE (HEuristic SAmpling based on VEctor feature). HESAVE introduces Iterator Vector to describe motion semantics of trajectory points. Furthermore, Iterator Vector Information is proposed to quantify the information of trajectory points based on Iterator. Moreover, HESAVE adopts a data-driven window called Sliding Mode Window to split multi-mode trajectory into isolated process units and a priority queue for each unit to select trajectory points. Extensive experiments on GeoLife 1.3 dataset show that HESAVE can gain more reservation of trajectory’s temporal and positioning information after sampling under the same sampling ratio compared to SQUISH. In addition, HESAVE’s computation resource consumption is quite acceptable.