Abstract
Cell phone data has become a major source for the scientific community to analyze human mobility behaviour [4][6], with wide ranging applications like traffic estimation [3][5] or disaster relief [7]. A typical task in this context is to estimate motion trajectories based on sequences of locations in the antenna network. We examine the question: given that a mobile device is logged in to an antenna with known position and characteristics, what can be said about the probability distribution of the position of the device?
Theoretical studies [1][2] of the positioning error often assume knowledge of signal strength of one or more connected antennas. However, cell phone records are often given in CDR or similar format, where only the main connected antenna is recorded. We examine two datasets of trajectory data with nearly 1.8 million data points, where location was determined using a GPS logger. Based on this data, we give estimates on the distribution of the positioning
errors.
Originalsprache | Englisch |
---|---|
Publikationsstatus | Veröffentlicht - 2013 |
Veranstaltung | NetMob 2013 - Dauer: 1 Mai 2013 → 3 Mai 2013 |
Konferenz
Konferenz | NetMob 2013 |
---|---|
Zeitraum | 1/05/13 → 3/05/13 |
Research Field
- Ehemaliges Research Field - Mobility Systems