Accurate Map Matching Method for Mobile Phone Signaling Data Under Spatio-Temporal Uncertainty

Understanding human mobility has become a greater demand in recent years. Among them, how to extract people’s travel trajectories and reconstruct them accurately is crucial to explain people’s mobility. Due to its spatio-temporal uncertainty and sparsity, however, map matching based on mobile phone signaling data (MSD) remains a challenge. In this paper, we introduce an innovative and precise Map Matching Method tailored for MSD, leveraging an incremental Hidden Markov Model (HMM) algorithm. A series of effective modules are put forward to address several critical challenges associated with MSD, including handling drifting data, ping-pong sequences, spatio-temporal sparsity, location uncertainty, back-and-forth movements, and U-turn errors. Then Incremental HMM is applied to provide the optimal path on the digital map. To assess the effectiveness of our proposed model, we conducted a comprehensive comparison against state-of-the-art methods and conducted ablation experiments. The results highlight that our model outperforms existing approaches, achieving 82.9%, 93.6%, and 0.88 in precision, recall and F1 based on the distance measurement, respectively. The proposed method makes it possible to work on real-time map matching based on large-scale MSD. In the future, we will conduct the exploration of human mobility based on MSD in cities and congestion detection of network traffic states.