Multi-Objective Optimization Based Service Pre-Deployment in Edge Computing

Edge computing enhances service quality by deploying application services closer to the data source at the network edge, providing rapid responses to the increasing service demands. However, edge servers, dispersed across the network, have limited computational and storage capabilities, necessitating efficient service deployment strategies to collaboratively process service request chains. Existing deployment methods focus solely on the response time of user requests, neglecting edge server resource utilization, leading to suboptimal deployments. This paper proposes a pre-deployment method for microservice compositions, aiming to minimize the response time of user requests and the proportion of services processed by cloud servers. By combining NSGA-II with the simulated annealing algorithm, our approach performs multi-objective optimization. Experimental results on a real-world dataset demonstrate that our method effectively improves edge service quality compared to baselines.