An adaptive service deployment algorithm for cloud-edge collaborative system based on speedup weights

In the contemporary landscape of edge computing, the deployment of services with stringent real-time requirements on edge devices is increasingly prevalent. However, the challenge of designing an effective service deployment strategy that optimally leverages both cloud and edge resources to deliver high-quality services in production environments persists, primarily due to resource constraints in edge devices. To tackle this issue, we introduce an adaptive service deployment algorithm that utilizes speedup weights for cloud-edge collaborative environments (SWD-AD). First, by comparing the execution and communication times of tasks in the cloud and at the edge, the speedup weights are calculated, and a service deployment algorithm is designed that takes into account both the speedup weights and resource consumption weights. Then, during the cluster operation, information on the task processing for each service is collected and their cumulative speedup weights are calculated. Utilizing a dynamic service adjustment strategy based on these cumulative speedup ratio weights, services are migrated between the cloud and the edge. Our performance evaluation experiments reveal that this strategy notably reduces the average response time of tasks by 29.38 and 25.86% compared to Swarm and Kubernetes (K8s) algorithms, respectively.