A Hybrid Data Plane SDN Architecture Based on Cloud-Native for Segment Routing Over IPv6
SRv6 (Segment Routing over IPv6) serves as the cornerstone for next-generation internet solutions. In SRv6 networks, SDN architectures address critical challenges like multi-cloud SRv6 traffic engineering through deep integration with SRH parsing and Kubernetes CNI. Existing SDN implementations face two key limitations: 1) Inability to concurrently configure hybrid data planes combining VPP and Linux nodes, which diminishes SRv6’s cross-plane advantages; 2) Operational complexity from maintaining synchronization between disparate controllers, hindering unified southbound management of SRv6 nodes. To resolve these issues, we propose a redesigned SDN architecture for cloud-native networks featuring two innovations: 1) A Linux SR Plugin (LSP) southbound interface enabling coordinated SRv6 support across heterogeneous nodes; 2) Enhanced northbound API capabilities through key-value store integration with cloud platforms. The implemented LSP plugin allows transparent management of hybrid data planes via business-focused API operations. Our experimental validation demonstrates that the LSP successfully verifies all SRv6 configurations supported by both the Linux and VPP data planes while achieving ≥400 configurations/second throughput. These results confirm the LSP’s capability to deliver unified southbound control for SRv6 across heterogeneous network environments.
Authors
- Chi Jiang (ORCID: https://orcid.org/0000-0002-9296-934X)
- Yuli Zhao (ORCID: https://orcid.org/0000-0001-7298-7463)
- Zhiliang Zhu (ORCID: https://orcid.org/0000-0002-6222-0864)
- Hai Yu (ORCID: https://orcid.org/0000-0002-8024-1781)
- D. Yuan (ORCID: https://orcid.org/0009-0008-3539-3429)
Institutions
- Shenyang Institute of Engineering (CN)
- CCTEG Shenyang Research Institute (CN)
- Northeastern University (CN)
Publication Details
- Journal
- IEEE Transactions on Network and Service Management
- Published
- 2025-07-14
- DOI
- https://doi.org/10.1109/tnsm.2025.3588535
- Primary Topic
- Software-Defined Networks and 5G
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- National Natural Science Foundation of China
- Higher Education Discipline Innovation Project