Indoor Mapping Structure Based on Cloud Platform for Seamless and Effective Indoor Localization

Abstract

Unlike outdoor spaces, indoor spaces are not naturally interconnected. To provide efficient indoor location-based services, we structured indoor spaces into Sector, Building, Level, and Spot, and connected them through a relational database on the Cloud Platform. The relational database enables the seamless interconnection, easy management and modification of each space. Additionally, engine systems for indoor localization are established on the Cloud Platform to provide accurate spatial and positioning information. These engines are designed to be independent, allowing efficient resource utilization based on users' service demands. We conducted successful and seamless test at COEX, one of the largest underground parking lots in Seoul, South Korea, demonstrating a location error of 4.795m and a floor estimation accuracy of 97.926% for seamless floor transitions. The approach presented in this paper enables efficient management of indoor spaces and provides seamless location services tailored to users' needs.