Data reduction in fog computing and internet of things: A systematic literature survey

Cloud computing, the most crucial computing and storage tool in IoT (Internet of Things), still meets various challenges. The remoteness of IoT end devices from cloud platforms may lead to significant issues for real-time applications such as disaster handling, healthcare application, etc. In order to address these issues, fog computing is a new platform with specialized features that perform essential IoT data management and real-time application management tasks. Data management in IoT via fog computing is critical for decreasing latency in real-time IoT applications and is needed to produce more professional knowledge and smart decisions. Reduction in the size of data sent to the cloud layer is a fundamental topic in data management on the fog computing platform. In this paper, to select and survey studies about data size reduction in fog computing, we applied the Systematic Literature Review (SLR) process to comprehend and classify the different topics and related approaches in this field. In addition, the studies presented in the edge computing field, which were close to our goal, were also investigated. The primary purpose of this study is to classify and analyze the fog data reduction (FDR) studies published between 2016 and 2022. The topics and related approaches of selected papers are presented in an approach-based taxonomy. The topics of provided taxonomy include data filtering, data compression, data aggregation, data prediction, data pattern recognition, and general facets and related approaches. Finally, open issues for FDR and relevant main challenges are presented in upcoming research.

Automated summary

This paper highlights the importance of cloud computing and fog computing in the Internet of Things, the critical role of data management, and the essential techniques for data size reduction in fog computing. The study focuses on classifying and analyzing FDR studies from 2016 to 2022, presenting relevant topics and methods, and identifying open issues and challenges for future research.