Motivation
Satellite Navigation (GNSS, Global Navigation Satellite Systems) is expected to become a game-changer for railway traffic management systems in the future, because with the support of GNSS train traffic can be management based on the continuously reported train positions instead of the fixed track-sections occupied by the trains. Due to the high safety-criticality of this application the GNSS data must fulfill very stringent quality criteria in the terms of accuracy, availability, integrity, etc. and must be robust against interferences from other radio sources and/or intentional jamming- and spoofing-attacks. The Swiss Federal Railway (SBB) operates a measurement wagon, which collects regularly GNSS data throughout the Swiss railway network. This data shall be analyzed with the support of AI methods to identify and characterize malfunctions caused by interferences, jamming and spoofing and assess the GNSS performance in various geographical environments.
 

Workpackages

1. Definition of relevant GNSS specific KPIs for

• Detection of malfunctions
• Characterization of malfunctions
• Interpretation of malfunctions (correlation with specific areas, times, etc.)
• Assessment of GNSS performance in specific geographical environments

2. Selection of appropriate AI methods and algorithms

3. Data selection and import

4. Data Analysis

• Without AI methods
• With AI methods
• White Listing of railway lines (good reception and no malfunctions)

5. Assessment and presentation of results

• GNSS performance evaluation in different environments
• Areas of GNSS Interference (jamming, spoofing)

Technical Specifications
Data Sources

• Sensor data (SBB LocLab)
• Map data (SBB LocLab)
• Terrain and 3-D building Data from Swiss Topo (publicly available)

Software/IT

• Python
• SQL