I<SUP>2</SUP>CLOUDMAN: DQN-Enhanced ILP Model for Optimal Virtual Machine Placement and Migration in Cloud Data Centers

Dogan O. D., Mert N. K., SCHMİDT K. V., SCHMİDT K. V.

33rd Conference on Signal Processing and Communications Applications-SIU-Annual, İstanbul, Turkey, 25 - 28 June 2025, (Full Text)

  • Publication Type: Conference Paper / Full Text
  • Doi Number: 10.1109/siu66497.2025.11111900
  • City: İstanbul
  • Country: Turkey
  • Middle East Technical University Affiliated: Yes

Abstract

Efficient resource allocation in cloud data centers (CDCs) for Infrastructure as a Service (IaaS) remains a critical challenge due to fluctuating workloads, power consumption constraints, and the need to minimize Service Level Agreement (SLA) violations. This paper proposes I(2)CLOUDMAN (Intelligent ILP CLOUD MANagement), a novel approach combining Integer Linear Programming (ILP) with a Deep Q-Network (DQN) to optimize virtual machine (VM) creat and migration. The ILP model minimizes power consumption while maximizing resource utilization, and the DQN dynamically adjusts the critical ILP parameters based on real-time workload conditions. This hybrid solution offers a scalable and adaptive framework for modern CDCs, effectively balancing operational efficiency and SLA compliance.