研究目的
Investigating the development of a novel SDN-based QoS control framework for industrial networks that achieves hard real-time QoS guarantees with high bandwidth utilization.
研究成果
The proposed function split QoS framework makes very quick runtime admission decisions while providing deterministic worst-case QoS guarantees. It achieves up to approximately 93% average link utilization in an example industrial communication scenario, coming close to the performance of a monolithic MIP that jointly optimizes routing and resource allocation. Future research directions include exploring a wider range of routing cost functions and scheduling policies, refining the background resource allocation, and considering statistical (soft) real-time QoS.
研究不足
The study focuses on industrial networks and may not be directly applicable to other types of networks. The computational complexity of the monolithic MIP approach is prohibitive for large networks, highlighting the need for the proposed function split approach.
1:Experimental Design and Method Selection:
The study employs a novel SDN-based QoS control framework that avoids a control loop over forwarding and control planes by maintaining an accurate network model through network calculus. It introduces a function split between delay-constrained least-cost routing on a network of queue links and resource allocation to these queue links.
2:Sample Selection and Data Sources:
The evaluation is based on simulations comparing the proposed function split approach with a monolithic mixed-integer programming (MIP)-based approach for joint routing and resource allocation.
3:List of Experimental Equipment and Materials:
The study utilizes software-defined networking (SDN) and network calculus for modeling and simulation purposes.
4:Experimental Procedures and Operational Workflow:
The proposed framework is evaluated through simulations that assess its ability to provide online runtime admission control and achieve high bandwidth utilization while meeting deterministic real-time QoS requirements.
5:Data Analysis Methods:
The performance of the proposed framework is analyzed in terms of bandwidth utilization and the ability to meet real-time QoS requirements, comparing it with the MIP-based approach.
独家科研数据包,助您复现前沿成果,加速创新突破
获取完整内容