A Trajectory Collaboration Based Map Matching Approach for Low-Sampling-Rate GPS Trajectories.

Area of Science:

• Geomatics Engineering
• Computer Science
• Data Science

Background:

• Global Positioning System (GPS) trajectories with low sampling rates present significant challenges for accurate map matching due to inherent uncertainty between data points.
• Existing map matching algorithms struggle to effectively process sparse GPS data, leading to reduced accuracy in various location-based applications.

Purpose of the Study:

• To propose and evaluate a novel collaborative map matching (CMM) method designed to enhance the accuracy of low-sampling-rate GPS trajectories.
• To address the limitations of current methods in handling uncertainty and missing information in sparse GPS data.

Main Methods:

• The proposed Collaborative Map Matching (CMM) method processes GPS trajectories in batches.
• It involves clustering similar GPS trajectories, resampling to supplement missing information, and extracting a collaborative trajectory for each cluster.
• Map matching is performed using the Longest Common Subsequence (LCSS) distance algorithm.

Main Results:

• Experimental results on both real and simulated datasets demonstrate that CMM significantly outperforms baseline methods.
• The proposed method shows superior effectiveness in matching low-sampling-rate GPS trajectories to the road network.
• CMM also exhibits improved efficiency compared to existing approaches.

Conclusions:

• The collaborative map matching (CMM) method offers a robust solution for improving the accuracy and efficiency of Global Positioning System (GPS) trajectory analysis with low sampling rates.
• CMM effectively mitigates the challenges posed by sparse GPS data, paving the way for more reliable location-based services.