Decentralized Edge-to-Cloud Load Balancing: Service Placement for the Internet of Things

The Internet of Things (IoT) requires a new processing paradigm that inherits the scalability cloud while minimizing network latency using resources closer to edge. Building up such flexibility within edge-to-cloud continuum consisting distributed networked ecosystem heterogeneous computing is challenging. Furthermore, IoT traffic dynamics and rising demand for low-latency services foster need response time balanced service placement. Load-balancing fog becomes cornerstone cost-effective system management operations. This paper studies two optimization objectives formulates decentralized load-balancing problem placement: (global) workload balance (local) quality (QoS), in terms cost deadline violation, deployment, unhosted services. proposed solution, EPOS Fog, introduces multi-agent collective learning utilizes nodes jointly input across minimize costs involved execution. agents locally generate possible assignments requests then cooperatively select an assignment their combination maximizes edge utilization minimizes execution cost. Extensive experimental evaluation with realistic Google cluster workloads on various networks demonstrates superior performance Fog QoS, compared approaches as First Fit exclusively Cloud-based. results confirm reduces delay 25% load-balance 90%. findings also demonstrate how computational can be utilized more cost-effectively by harvesting intelligence.