A Structured Framework for Selecting Cloud or Fog Computing in Application-Specific Contexts

This research work proposes a structured, four-stage framework to guide the selection between Cloud and Fog computing architectures based on specific application requirements. Experimental evaluations conducted on IoT-based smart city systems and autonomous vehicle simulations reveal that Fog Computing reduces latency to 10–50 ms and decreases bandwidth consumption by approximately 45% compared to Cloud Computing. Cloud Computing demonstrates superior performance in handling large-scale batch processing with high storage capacity but incurs higher latency (250–600 ms). Security and resource management trade-offs are also analysed, showing Fog’s advantage in localized processing but increased complexity in device coordination. The framework integrates these findings into a weighted evaluation matrix incorporating performance, security, cost, and scalability metrics. Practical case studies validate the framework’s effectiveness in recommending optimal computing paradigms for diverse industry scenarios including healthcare, smart cities, and industrial IoT.