Zero-touch network and service management in 6G
Monday, 3 June 2024, 09:00-12:30, Room Gorilla Room 4
Speaker:
- Johann M. Marquez-Barja (Univ. Antwerp – imec, IDLab – Faculty of Applied Engineering, BE)
- Nina Slamnik-Kriještorac (Univ. Antwerp – imec, IDLab – Faculty of Applied Engineering, BE)
- Xhulio Limani (Univ. Antwerp – imec, IDLab – Faculty of Applied Engineering, BE)
Motivation and Context
Orchestrating 6G services that tailored for diverse vertical industries (automotive, eHealth, transport & logistics, etc.), demands innovative strategies to overcome limitations inherent in existing management techniques. These limitations can lead to significant delays in network management operations, adversely affecting service performance. For example, traditional approaches like the human-in-the-loop method are slow and error-prone, while closed-loop control using rule-based algorithms is challenging due to the intricate configuration of numerous network and service parameters. A promising solution based on Zero-touch Network and Service Management (ZSM) involves leveraging Artificial Intelligence (AI) and Machine Learning (ML) in conjunction with Network Function Virtualization (NFV) and Software Defined Networking (SDN) to introduce automation and intelligence, optimizing orchestration operations. In the context of 6G, enhancing the performance of vertical services can be achieved through cloud-based services and Multi-Access Edge Computing (MEC) units located closer to users. In particular, the growing demand for connected vehicles requires efficient distribution of network and computing resources to ensure the required levels of End-to-End (E2E) latency, throughput, and reliability, crucial for traffic participants’ safety. We have been focused on developing a ZSM framework for vertical services using the Smart Highway testbed, aiming to enhance and validate vertical service performance in a real-world setting. The proposed framework involves decision-making processes (zero touch services) aligned with ETSI ZSM principles and intent-driven management. The decisions are based on the workload and availability of MEC units, aligning with the demands of vertical services and the mobility of connected users. One of our recent efforts includes the implementation of a Deep Reinforcement Learning (DRL)-based service to optimize decision-making, compared to a simpler rule-based algorithm. Results demonstrate that DRL swiftly adapts to the dynamic testbed environment, surpassing conventional rule-based approaches. This suggests that the DRL algorithm can enhance the decision-making process, ultimately reducing E2E latency for vertical industries.
Structure and Content
The proposed tutorial will take half a day (3 hours). The tentative outline is presented below.
Part A – Theoretical concepts of Zero-touch network and service management in 6G (100 minutes; led by Prof. Johann M. Marquez-Barja, and Dr. Nina Slamnik-Kriještorac):
- Introduction: network and service management in previous Gs and where are we now; ETSI standardization activities; relation between ZSM and Intent-based network management
- Foundational technologies: A short trip through Network Function Virtualization (NFV), Software Defined Networking (SDN), and Artificial Intelligence (AI)/Machine Learning (ML), and how they blend together to improve network and service management techniques
- Timeline of Zero-touch Network and Service management: Initiatives before ETSI ISG ZSM (founded in 2017), and progress since then
- ZSM operations: Orchestration, Control, and Monitoring
- Zero-touch services in 6G ecosystem: Analysis of AI/ML techniques for realizing relevant decision-making processes in 6G operations; Benefits, challenges, and implications of running zero-touch services and network intelligence at the network edge
- 6G Applications: Vertical services tailored to address demands and needs from relevant 6G verticals; Standardized application frameworks; Network
awareness, energy-awareness, and security-awareness of 6G Applications - Relevant performance metrics and validation methodologies: Measuring the impact of ZSM decisions on network and service performance, relevant Key
Performance Indicators (KPIs) and Key Value Indicators (KVIs) - ZSM research and implementation in European projects: The examples, Proof-of-Concepts, trial implementations
Coffee break 15 minutes
Part B – Building Proof-of-Concepts for ZSM in real-life settings for 6G verticals (65 minutes; led by Dr. Nina Slamnik-Kriještorac and Xhulio Limani)
- Implementation of ZSM: Software and hardware requirements; examples; frequently used software frameworks for management and orchestration
- ZSM for vertical services: Positioning of vertical services in ZSM ecosystem in 6G; implementation practices; interfaces between applications, orchestrators, controllers, and monitoring entities, to achieve the full potential of network/energy/security-awareness of applications
- Examples of 6G applications using ZSM: Vertical service for smart traffic management use case from the TrialsNet project; a high-level overview of the system architecture is presented, including the details on service operation, and hints on how this type of service can be managed and orchestrated in distributed MEC environments
- Building a Proof-of-Concept on the testbeds: We describe some useful practices that we adopted to build a proof-of concept for testing and validating the impact of ZSM on vehicular services, while utilizing the real-life testbeds such as Smart Highway and Open5G in Antwerp, Belgium
- Live experimentation: Small-scale experiment showing performance evaluation and metric collection in real-time for two zero-touch services helping edge orchestrators to make efficient decisions
The content and material of this tutorial is backed by knowledge and experience of the Flexible Networking Group at IDLab, imec research group at the University of Antwerp. The group is led by Prof. Marquez-Barja and Dr. Slamnik-Kriještorac, who built the expertise on the proposed topics by working of numerous European and national projects in the context of advanced networking for vertical industries. The referenced papers provide both the background and the practice for this tutorial, transitioning from techniques to share network resources for 5G to orchestrating services at the MEC to fulfil critical services for emergency vehicles.
