Automated, Dynamic, and QoS-Aware Microservice Composition: A Systematic Literature Review

Morteza Khani Dehnoi¹

1) Department of Computer Engineering, Faculty of Electrical and Computer Engineering, Quchan University of Technology, Quchan, Iran

IOI Code : ABEB-ACABCH
Publication : 4th International Conference on Recent Advances in Engineering, Innovation and Technology - Belgium(eitconf.com)
Abstract :
    Microservices have become the backbone of many scalable software systems. However, when composing these services into larger capabilities, two problems quickly become complicated: performing the composition automatically without human intervention, and preserving Quality of Service (QoS) under conditions where service status changes continuously. This paper presents a systematic review of existing research in this area. Through searches in reputable databases and after screening, key studies were categorized into three technical strands: runtime decision-making with reinforcement learning, composition optimization with metaheuristic and graph-based methods, and MAPE-K feedback loops for adaptive reconfiguration. The main finding is that all three strands share a common weakness: none can simultaneously model state-dependent QoS, respect coupling constraints between services, and make decisions locally while being aware of rebinding costs. This gap defines the path for future research. The review further proposes four directions: microservice-native problem modeling, hybrid learning and optimization architectures, standardized evaluation benchmarks, and integration of security and trust constraints at the composition level.

Keywords : microservice composition qos aware dynamic recomposition systematic literature review runtime adaptation