.png){kind=icon}
返回
CoCo: Quantifying Correlations between Mobility Traces using Sensor Data from Smartphones
.png){kind=icon}
OA
AI
DOI：10.1145/3457139.png)
.png){kind=icon}
原文求助.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
As mobility is an important key to many applications, this work proposes a location-less model to represent mobility that is used to quantify correlations between mobility traces collected by built-in sensors on smartphones. We analyze the mobility correlations from two aspects: co-direction relationship and co-movement relationship. The former is to quantify the similarity of macroscopic moving directions between mobility traces, whereas the latter is to quantify the similarity of their microscopic vibrations. To verify the merits of the two proposed metrics, an exemplary use case, termed co-mobility detection, is considered to determine if two mobile devices share the same journey on the same mobile entity (e.g., carried by the same person). Comprehensive experiments with diverse combinations of mobility traces are conducted in three different environments with different density of Wi-Fi networks. The experimental results indicate that the proposed metrics can effectively evaluate both the coarse-grained similarity of moving directions and the fine-grained similarity of movement variations along mobility traces. The accuracy of the co-mobility detection algorithm can achieve 90% on average for mobility traces with a duration of 70 s.
Keyword:
Cyber-physical systems
mobile sensing
mobility analytics
smart cities
urban computing
期刊
IF:
4.1
论文数:
896
被引数:
1.9K
机构
引用论文
Indoor Location Learning Over Wireless Fingerprinting System With Particle Markov Chain Model.png)
IEEE ACCESS
OAAI
IEEE ACCESSIF3.6
OAAI分享收藏
查看更多内容