Journal of Communications and Networks · 2024 · 13 citations · 32 references
ComputingComputer SystemsInformation TechnologyComputer NetworksQueuing TheorySystems ArchitectureNetwork ArchitectureNetwork Performance AnalysisNetwork Traffic AnalysisRoutingService-oriented ArchitectureSoftware-defined NetworkingNetwork Traffic EngineeringPrioritizationQos ManagementServer SelectionService-level AgreementsSoftware-defined Traffic EngineeringUser SatisfactionUser Satisfaction Assessment
The future integration of software-defined network (SDN) with the service-oriented architecture (SOA) paradigm requires new solutions to ensure the quality of service (QoS) according to the users' requirements. The paper presents a user experience-centric approach to traffic engineering and QoS/quality of experience (QoE) support for service-oriented software-defined network (SOSDN) architecture. This approach is to enable end-to-end QoS across the networking and computing domain by monitoring and agreeing on the dynamic state of their functioning. The proposed SOSDN is based on improved traffic engineering techniques, such as adaptive prioritization of services, server selection, and QoS/QoE-based routing. The developed adaptive service prioritization algorithm automatically changes the priority of flows in the network operation mode by the SDN controller for individual users under the concluded service level agreements (SLA) contract. We proposed a mathematical model of correlation of user satisfaction level by QoE score with technical QoS parameters. This model is based on the normalized value of the integral additive QoS criterion. Accordingly, ensuring the ordered user-centric QoS/QoE is carried out by means of proposed multi-criteria adaptive routing of data flows, the metric of which is based on the integral additive QoS criterion. The simulation results showed that, in contrast to known practical solutions, the integrated use of the proposed method of adaptive multi-criteria routing and prioritization of data flows provides a high level of QoE required by users in the SOSDN paradigm.
32
DROM: Optimizing the Routing in Software-Defined Networks With Deep Reinforcement Learning
Changhe Yu, Julong Lan, Zehua Guo et al. · IEEE Access · 2018 · 184 citations · Full text