A PRISMA-Based Systematic Review of Cloud- Edge Orchestration Using the MAPE-K Framework

Although cloud computing offers abundant computational capacity, it suffers from inherent latency. Edge computing mitigates this by processing data closer to the edge of the infrastructure, thereby reducing latency and improving performance. However, challenges arise owing to inadequate edge processing capacity and significant cloud latency. A viable solution to address these challenges is cloud-edge integration, a collaborative resource distribution model. This study examines cloud-edge orchestration, focusing on performance and efficiency improvements through orchestration techniques. We performed a comprehensive systematic literature review using the PRISMA model, which compiled 10,389 records from the decade spanning 2015 to 2024, filtered down to 89 studies. Using the Monitor-Analyze-Plan-Execute over shared Knowledge (MAPE-K) framework, we assessed cloud-edge orchestration and categorized the performance metrics. Additionally, we evaluate the task distribution criteria between cloud and edge computing environments, identify challenges, and outline prospective directions. Our findings provide insights into optimizing cloud-edge computing for mainstream applications, ensuring improved resource management and efficiency in cloud-edge computing.