Toward an autonomous Smart Home: a three-layer Edge-Fog-Cloud architecture with latency analysis

Abstract

Due to the rapid evolution of smart devices, recently, the smart home (SH) sector has attracted attention and faced a fast expansion. However, the current cloud-based SH architecture and the ever-growth of the smart-things number and applications have brought several challenges to SH like the user-dependent automation tasks, the user-dependent SH management system, and the relatively large delay response time. These challenges create a gap between the actual SH and the envisioned autonomous SH that can control the SH objects independently and automate the home intelligently. Reducing this gap became the main target of the current SH research works. This paper builds on the network softwarization and virtualization techniques along with the edge, fog, and cloud computing to propose a three-layer Edge-Fog-Cloud SH (EFC-SH) intelligence architecture that can enable the autonomous SH. Then, it derives the optimal trade-off between the workload and the latency at each layer. Simulation results show that the proposed EFC-SH architecture achieves lower latency compared to the cloud-only or fog-only architectures.