Crowdsourced Reconstruction of Cellular Networks to Serve Outdoor Positioning: Modeling, Validation and Analysis

Area of Science:

• Computer Science
• Geomatics Engineering
• Telecommunications

Background:

• Outdoor fingerprinting relies on cellular tower signals for positioning.
• Existing crowdsourcing efforts lack focus on cellular network topology, crucial for accurate modeling.
• Network modeling is complex due to diverse technologies, spatio-temporal dynamics, and evolving configurations.

Purpose of the Study:

• To develop a system for synthesizing and modeling cellular network topology from crowdsourced signal fingerprints.
• To create a spatio-temporal database for managing heterogeneous cellular data and network evolution.
• To enable robust outdoor positioning by improving the understanding of cellular network structure.

Main Methods:

• Synthesizing conceptual and logical schemas of cellular networks from signal fingerprints.
• Building a relational spatio-temporal database for crowdsourced data collection and management.
• Developing methods to handle network configuration evolution, including cell renaming and measurement correction.

Main Results:

• Successfully synthesized detailed cellular network schemas from signal data.
• Created and populated a spatio-temporal database with heterogeneous cellular observations.
• Demonstrated the system's ability to manage network evolution and correct inconsistent measurements.
• Enabled comprehensive spatial and temporal analyses of cellular network arrangement and changes.

Conclusions:

• The developed system effectively reconstructs and maintains knowledge of cellular network topology using crowdsourced data.
• This approach significantly enhances the foundation for accurate outdoor positioning systems.
• The system provides a scalable solution for managing and analyzing dynamic cellular network information.