关键词: 物联网, 智能家电, 云–边–端协同, 边缘计算, 动态任务调度, 轻量化AI

Abstract: Against the background of the explosive growth of IoT smart home appliances, traditional cloud computing architectures face key challenges such as excessively high end-to-end latency, limited network bandwidth, and prominent privacy and security risks. A cloud-edge-end collaborative architecture tailored for IoT smart home appliance scenarios is proposed. The architecture adopts a layered design, clarifies the functional boundaries of the device perception layer, edge computing layer, and cloud service layer, and constructs a three-dimensional collaboration system of end-end, edge-end, and cloud-edge with the integration of "scenario-task-resource" three-driven collaborative scheduling logic. A cloud-edge-end dynamic task scheduling algorithm (DMOA) is developed to balance latency, energy consumption, and resource utilization through cross-layer resource perception, dynamic weight adjustment, and hierarchical task allocation. Experimental verification shows that compared with the traditional pure cloud architecture, the proposed architecture reduces data processing latency by 51.3%, decreases network bandwidth occupation by 43.7%, and improves system energy efficiency by 18.4%, providing technical support for the large-scale deployment and collaborative control of smart home appliances.

Key words: Internet of things, Smart home appliances, Cloud-edge-end collaboration, Edge computing, Dynamic task scheduling, Lightweight AI