User-Centric and Community-Based Microservices Placement for Energy Efficiency

The growth of IoT and connected devices has increased demand for low-latency, energy-efficient processing across the Cloud-Fog-Edge continuum. While microservices enable scalable distributed computing, their placement remains challenging due to dynamic resource needs and interdependencies. This work proposes a graph-based microservice placement approach using user-centered local community detection. By integrating user nodes, our method adapts to shifting demands and resource availability, reducing energy consumption and communication overhead. Additionally, strategic mutualization and controlled duplication further enhance efficiency while preserving response time and resource constraints. Our results highlight the effectiveness of user-centric strategies in achieving scalable and sustainable deployments, reducing energy consumption by approximately 50% compared to state-of-the-art global methods while slightly improving deployment time.