Quantum-Classical Weighted Clique Detection for Microservice Deployment in Cloud Systems

Abstract

The widespread adoption of microservice architectures in large-scale cloud systems introduces challenges related to the optimal deployment of service components and the optimization of inter-service communication. This study proposes a quantum-classical hybrid approach to address this problem, which can be considered an NP-hard problem at large scale, with the aim of achieving an optimal deployment of microservices that exhibit intensive interactions. In the proposed method, microservice interactions are modeled as a weighted graph and transferred to a quantum environment for the purpose of maximum clique detection. Modeling the problem using a weighted graph, where microservice interaction parameters such as latency, communication frequency, and data volume are evaluated as composite edge weights, enables the deployment problem to be grounded in system-oriented criteria and treated as a multi-dimensional optimization task. In this study, all valid clique structures are correctly identified, thereby enabling optimization strategies to be executed within the quantum computing environment. © 2026 IEEE.