A survey on data fusion approaches in IoT-based smart cities: Smart applications, taxonomies, challenges, and future research directions

Abstract

Rapidly increasing urbanization leads to the need for more comfortable and reliable living spaces. The smart city paradigm needs to be renewed daily to provide smarter solutions to citizens' needs and problems and ensure sustainable living. The Internet of Things is one of the most widely used smart city methodologies. The Internet of Things (IoT) aims to enable objects in the physical world and cyberspace to communicate. IoT technology produces a very large amount of raw and generally heterogeneous data. Data fusion techniques are gaining popularity to manage this large amount of data. With data fusion methods, smarter structures can be built from this raw data to reduce data size, optimize traffic, and extract useful information. This study provides a comprehensive perspective and opportunities for different areas of IoT applications of smart city systems. In addition to providing a detailed taxonomy framework for data fusion levels in various criteria, the seven-layer IoT architecture implemented in smart cities is discussed. Finally, the paper concludes by mentioning the difficulties encountered in applying data fusion methods, proposing solutions to these difficulties, and presenting future work trends based on the studies conducted.