Check my Street Credibility

Art der Arbeit:

Studien-/Bachelor-/Diplom-/Masterarbeit

Betreuer:

Hammer, Andreas
Lehrstuhl für Informatik 1 (IT-Sicherheitsinfrastrukturen)
Telefon 09131/85-69596, Fax 09131/85-69919, E-Mail: andreas.hammer@fau.de

Beschreibung der Arbeit:

Gangster rappers and geolocation traces in court have one thing in common: Questionable street credibility. We focus on discussing the latter here. In a (technically) perfect world, all GPS traces created by phones, smart watches or cars would create perfectly accurate and precise mappings of time to location of a suspect. In reality, bad reception and generally bad quality lead to unprecise data that is unfit to be used directly as evidence due to noise. A question arises: Is this location trace plausible? Does speed, heading and general movement check out with respect to the real environment?

This thesis aims to discuss this question with the help of OpenStreetMap data. The OSM project provides not only rendered map tiles but also raw annotated map data that is searchable and analyzeable. For the purpose of evidence enrichment, raw GPS traces will be snapped to "real" streets and deviations will be made visible. Walking on a sidewalk is not interesting. Suddenly stopping and deviating into the forest may be. Later returning to the sidewalk and continuing with a higher speed is even more valuable of an information.

In this thesis, you will:

• Get familiar with GPS traces, file formats and the OpenStreetMap project

• Discuss data from OpenStreetMap that may be valuable for forensic purposes

• Discuss trustworthyness of traces, map data and the connection of the two

• Build a system to snap traces to street networks and calculate deviations

• Automatically generate a report that lists and explains anomalies and phenomena

• Evaluate this method and discuss limitiations
  

It is up to you whether you want to work on this topic in English or German.

Vorausgesetzte Vorlesungen bzw. Kenntnisse:

Required:

• At least one lecture of our lab

• Basic skills in Python or a comparable language
  

Recommended:

• Lecture: "AppITSec" or "SecSys"

• Lecture: "ForensInfI"

• Lecture: "ForensInfII"

• Basic understanding of geospatial data and OpenStreetMap

Bearbeitungszustand:

Die Arbeit ist bereits abgeschlossen.