Skip to Main content Skip to Navigation
Theses

Cloud-native optical network automation platforms

van Quan Pham 1, 2, 3 
Abstract : Optical communication management and control are transforming to integrate new capabilities such as intent-based network management, closed-loop control automation, and multi-stakeholder orchestration. These capabilities are driven by the new connectivity requirements between data centers to enable future generations of services: Beyond 5G (B5G) and 6G applications offered at the edges of optical networks. The next generation of optical network management and control architectures will entail Software-Defined Networking (SDN) principles for the disaggregation of future optical systems. The current optical network controllers and managers are intrinsically proprietary and, consequently, restricted in openness, scalability, and flexibility. This Ph.D. thesis investigates and proposes breakthrough software architectures with: (i) their control functions for the optical systems and (ii) their management functions for optical connection services of Open Disaggregated Optical Networks. After explaining SDN architectures in the context and the constraints of optical switching and transmission networks, the thesis explains the challenges of current optical networks to transition towards the control of Partially Disaggregated Optical Networks as a first step and the control of Fully Disaggregated Optical Networks as the ultimate step. Novel software-defined optical network automation platforms with control functions based on micro-services are described pragmatically, considering open-source software frameworks and several open forums providing their languages and their data models. Their protocols are being developed for devices, network topology, and communication services. Next, the thesis described how control functions are designed as cloud-native network functions (CNF), enabling continuous integration and continuous development of cloud-native optical networking platforms. Automated optical channel path computation functions as services a re first addressed. These optical channel path computation services are described by explaining how the routing constraints defined by the evolutions of optical system capabilities can be integrated into the path computation engines (PCE). Several PCE algorithms for optical channel routing and spectrum allocation are presented, and their performances are compared in terms of reasonable or possibly optimal spectrum allocation.Subsequent to the concepts of automated optical channel path computation functions as services, the thesis proposes automated optical channel defragmentation functions as services to re-arrange the placements of optical channels for better and possibly optimal use of the spectrum grid to gain resources.From the evaluations of these different container-based optical control functions, several optical channel control automation scenarios are described to prove their concepts using a network bench in a lab and demonstrate the potential applications of optical CNFs.Finally, the thesis concludes with the synthesis of these research works and the future challenges to make the control and the management of optical networks more unified and streamlined to enable optical communications to be designed and an asset as connectivity services for future generation services.
Complete list of metadata

https://tel.archives-ouvertes.fr/tel-03702081
Contributor : ABES STAR :  Contact
Submitted on : Wednesday, June 22, 2022 - 5:15:11 PM
Last modification on : Friday, June 24, 2022 - 3:05:38 AM
Long-term archiving on: : Friday, September 23, 2022 - 7:04:16 PM

File

78378Pham_2022.pdf
Version validated by the jury (STAR)

Identifiers

  • HAL Id : tel-03702081, version 1

Citation

van Quan Pham. Cloud-native optical network automation platforms. Networking and Internet Architecture [cs.NI]. Institut Polytechnique de Paris, 2022. English. ⟨NNT : 2022IPPAS005⟩. ⟨tel-03702081⟩

Share

Metrics

Record views

77

Files downloads

10