Edge-to-Cloud Continuum: Optimizing Latency and Scalability in Next-Generation Applications

Abstract

The edge-to-cloud continuum represents a hybrid computing architecture combining local edge resources with centralized cloud infrastructure, designed to meet the demanding latency and scalability needs of nextgeneration applications—such as IoT, AR/VR, real-time analytics, and autonomous systems. This paper performs a comprehensive review of the continuum, analyzing how edge and cloud resources can be orchestrated for optimal performance. We surveyed literature on fog and edge computing paradigms, resource orchestration, and federated execution models. Utilizing simulation studies and architectural surveys, we evaluate key performance metrics— including latency, bandwidth, resource utilization, and scalability (e.g., via IoTSim-Osmosis and RECAP frameworks). Findings show that deploying compute and storage closer to end devices significantly reduces latency, while centralized resources retain scalability and consolidation benefits. Models like Opportunistic Edge Computing (OEC) and hierarchical IoT–Fog–Cloud architectures support dynamic resource sharing and localized processing but face challenges in heterogeneity and orchestration. We propose a generic deployment workflow: (1) classify application requirements; (2) identify edge vs. cloud tasks; (3) design resource provisioning and caching strategies; (4) simulate/benchmark using continuum-aware platforms; (5) deploy with orchestration tools; and (6) monitor and adjust based on QoS feedback. Advantages of this continuum approach include ultra-low latency, improved bandwidth efficiency, and enhanced scalability; disadvantages involve increased system complexity, heterogeneous resource management, and security concerns in distributed nodes. In conclusion, the edge-to-cloud continuum is essential to support real-time, scalable applications. Future work should explore AI-driven resource orchestration, edge-native serverless platforms, and seamless interoperability across heterogeneous continuum layers.