Start from Scratch: A Crowdsourcing-Based Data Fusion Approach to Support Location-Aware Applications

Area of Science:

• Computer Science
• Electrical Engineering
• Geomatics Engineering

Background:

• Indoor location-based technology is crucial for modern transportation and the Internet of Things (IoT).
• Crowdsourcing is increasingly used to improve indoor localization accuracy and efficiency.
• Challenges exist in fusing heterogeneous crowdsourced data due to variations in collection parameters (users, locations, time, noise).

Purpose of the Study:

• To address the limitations of existing data fusion methods in crowdsourced indoor localization.
• To propose a general solution for fusing multi-dimensional crowdsourced data.
• To enhance the quality of crowdsourcing services by aligning data with consistent time and location stamps.

Main Methods:

• A novel data fusion approach utilizing only sensory data features.
• Alignment of multi-dimensional crowdsourced data with consistent time and location stamps.
• Development of a method to build high-quality crowdsourcing services from raw data.

Main Results:

• The proposed method effectively fuses multi-dimensional crowdsourced data.
• Consistent time and location stamping is achieved using sensory data features.
• Experimental evaluations validate the effectiveness of the proposed fusion technique.

Conclusions:

• The developed method offers a more general and effective solution for crowdsourced indoor localization.
• Leveraging multi-dimensional sensory data features significantly improves data fusion and localization quality.
• The approach has the potential to enhance various crowdsourcing services reliant on accurate indoor positioning.