Sherif Zahran

Internet of Things (IoT) sensors produce a huge amount of data every second. This unexpected data volume needs significant effort to manage, structure, save, and process to produce effective information and best utilize it. Extracting valuable knowledge from this heterogonous data is a high demanding need. Making accurate decisions based on this data was a dream in the near past. Data management and handling data heterogeneity locally on sensor boards and regionally on local servers using Fog and centrally on Cloud acts as an enabler for giving IoT an excellent value even if it faces many complexities and challenges, as discussed in this paper. Moreover, data management in IoT is a serious issue due to communication among billions of devices, which generate massive datasets. Data analysis on such a big volume of data is a difficult undertaking due to the lack of any standard. A definition of IoT-based data should be established to determine what is available and how it can be used. A study like this also points to the need for innovative approaches to deal with such problems. It's a great difficulty to cope with such a range of data as IoT delivers processing nodes in the form of smart nodes due to the heterogeneity of linked nodes, varied data rates, and formats. The Industrial IoT involves real-time analysis of data from interconnected sensor devices. It provides a solid foundation for big data research. Real-time data streaming adds an extra dimension of complexity and involves some more challenges. we briefly outline current open issues of edge computing and cloud computing platforms based on our literature survey for IoT data streaming heterogeneity. Finally, we propose Fog of Things Framework as a solution for all these challenges.