Citywide road-network traffic monitoring using large-scale mobile signaling data
Qiuyang Huang,Yongjian Yang,Yuanbo Xu,Funing Yang,Zhilu Yuan,Yongxiong Sun
DOI: https://doi.org/10.1016/j.neucom.2020.07.150
IF: 6
2021-07-01
Neurocomputing
Abstract:Road-network traffic monitoring on city-scale is critical for a wide range of applications, such as traffic forecasting, congestion identification, traffic safety, and urban planning, etc. Despite the fruitful research outcomes, however, most traffic monitoring models suffered from limited coverage, data sparsity, and data deviation, which leads to a biased and inaccurate result. With the widespread usage of mobile phones, mobile signaling data is of great value for various fields, especially for monitoring urban traffic. Thousands cell towers are distributed in the urban area, which can serve as ubiquitous sensors. Specifically, a mobile phone will passively generate a mobile signaling record that contains users’ spatiotemporal information. When mobile phone users move with their phones, their phones will interact with cell towers and these towers can obtain their mobile signaling records. And these signaling records contain sufficient information for traffic monitoring. However, there also exists excessive noise in signaling records, which makes most monitoring models abandon these data. In this paper, we present the Urban-STM scheme, which utilizes large-scale anonymous and coarse-grained mobile signaling data to infer road-network traffic conditions. We apply our scheme to a real-world signaling dataset in Changchun city and present an extensive validation study based on 2000 taxicabs’ GPS trajectories. Experiment results show that our scheme improves traffic monitoring performance in terms of coverage and accuracy.
computer science, artificial intelligence