Supervised Learning for Optimal Latency-Aware Microservice Placement in Edge Computing Environments
Hirotaka Kasahara, Ryuichi Kitajima, Jun Towada, Osamu Sato · 2025
Edge computing (EC), which facilitates real-time data processing by computing near data source points, is crucial for latency sensitive applications requiring real-time and seamless data processing. A significant challenge in EC is optimal placement of applications to ensure that data processing is completed within a specified end-to-end (E2E) timeframe, considering both communication latency between computational nodes and processing time at each node. Existing container orchestration tools fall short of meeting these E2E latency requirements. This paper proposes a supervised learning-based method that jointly addresses latency and processing time for the optimal placement of applications in EC environments. The model's performance is compared with exact solutions, demonstrating that the proposed method effectively learns optimal placement policies for microservice-based applications. The proposed model surpasses the performance of conventional MLP, GAT, and GCN-based models, achieving an accuracy of 94.8% on a validation dataset. The difference between the inferred and exact solutions is less than 1.6% compared in the mean objective function. The findings of this study illustrate the effectiveness of the proposed learning-based method for microservice-based applications.