CRISP-DM User Mobility Determined IoT PlacementWithin a Real-World Smart Building for Resource Efficient Fog Computing Environments

The introduction of Fog computing to the traditional Cloud computing architecture addresses the limitations of Cloud computing by optimising resource efficiencies such as energy consumption, latency, and network bandwidth utilisation. This research focuses on Fog computing in a smart building environment where heterogeneous Internet of Things (IoT) devices and sensors are placed in a standard way across the smart building. Research trends indicate that Fog computing optimises resource efficiency. However, this study determines how resource efficiency can be optimised using physical and spatial methods through enhanced smart IoT devices and sensor placements. CRISP-DM methodology was selected for its standardised format and 6-step iterative process to validate this study in a real-world smart building, with real-world smart IoT devices and sensors generating real-world data usage. IFogSim2 network simulation toolkit also supports and validates this study. The evaluation and analysis of the results demonstrated that the data management's impact on resources was further optimised through physical and spatial adjustments of smart IoT devices and sensor placements. The results indicate that using multiple heterogeneous metrics simultaneously and real-world data achieved a maximum percentage saving of 1.66% and 39% for energy consumption and network bandwidth utilisation. In comparison, latency was 88.51% savings for CCTV data management, including user GPS movement data, and enhanced and optimised resource efficiency. On that basis, it is recommended that smart building design and planning adopt this method to improve resource efficiency further in a smart building and Fog computing environment. The limitations of this study are limited user sample movements.