The evolution toward AI-native networks necessitates seamless Artificial Intelligence (AI) integration, which demands privacy-preserving data management, scalable model deployment, and optimized resource allocation. To address these challenges, we introduce key AI enablers as Data Operations (DataOps), Machine Learning Operations (MLOps), and AI as a Service (AIaaS). DataOps enables secure, privacy-preserving data collection through techniques such as Differential Privacy (DP) and secure aggregation, facilitating AI-driven insights without exposing sensitive user data. MLOps enhances the AI life cycle by leveraging distributed learning paradigms, including horizontal and Vertical Federated Learning (VFL), supported by Split Learning (SL). AIaaS extends these capabilities by exposing AI models and services through standardized APIs, enabling on-demand training, inference, and automation. By integrating AIaaS with DataOps and MLOps, networks can achieve greater intelligence, adaptability, and compliance with privacy and interoperability standards. This paper introduces a coherent architectural framework and operational strategy for embedding AI-driven intelligence into 6G networks, with a focus on key innovations in data governance, AI model coordination, and service exposure.

Cloud-native rethinks the application architecture by embracing a micro-service approach, where each microservice is packaged into containers to run in a centralized or an edge cloud. When deploying the container running the micro-service, the tenant has to specify the amount of CPU and memory limit to run their workload. However, it is not straightforward for a tenant to know in advance the computing amount that allows running the microservice optimally. This will impact the service performances and the infrastructure provider, particularly if the resource overprovisioning approach is used. To overcome this issue, we conduct in this paper an experimental study aiming to detect if a tenant’s configuration allows running its service optimally. We run several experiments on a cloud-native platform, using different types of applications under different resource configurations. The obtained results provide insights on how to detect and correct performance degradation due to misconfiguration of the service resource.

Service Function Chaining (SFC) is a networking concept by which traffic is steered through a set of ordered functions composing an end-to-end service. It represents one of the facilitating technologies for 5G, and is enabled by the Network Function Virtualization (NFV) and Software Defined Networks (SDN) paradigms. In the multi-domain context, SFC placement faces new challenges related to the lack of visibility on the local domain's networks. Indeed, the domain operators are often reluctant to unveil details on their topology to external parties. Furthermore, the new 5G use cases introduce new requirements for services such as end-to-end latency, and a minimal guaranteed bandwidth that the placement process needs to optimize simultaneously. In this article, we propose a centralized framework that allows SFC partitioning and embedding over multiple domains with a limited visibility over the global infrastructure. We model the multi-objective SFC placement problem using the Physical Programming method, which allows the expression of the Decision Maker's preferences through meaningful parameters, and propose an exact algorithm as well as a scalable heuristic solution. We then perform an extensive evaluation of the framework as well as the proposed algorithms. The results demonstrate our solution's effectiveness with a limited visibility on the network.

Service Function Chaining (SFC) has emerged as a promising technology for 5G and beyond. It leverages Network Function Virtualization (NFV) and Software Defined Networking (SDN) and allows the decomposition of a given service into a set of blocks that successively process data. The SFC placement issue has been extensively studied in the literature, and different solutions have been proposed using mathematical models and heuristics. More recently, Reinforcement Learning (RL) has emerged as a tool for decision-making that allows agents to elaborate policies based on the environment's feedback. In this paper, we study the benefits of using Deep Reinforcement Learning methods for the multi-domain SFC placement problem. We propose a Deep Deterministic Policy Gradient (DDPG) approach, where Linear Physical Programming is employed to generate rewards that reflect the solution's quality in terms of cost and latency. Through our experiments, we are able to demonstrate the efficiency of our approach with results that satisfy the SLA requirements.

Service Function Chaining (SFC) is the process of decomposing a network service into multiple functions that successively process packets to deliver the end-to-end service. In a multi-domain context, SFC placement is a challenging problem due to limited knowledge of the infrastructure of the local domains, which complicates the process of finding the optimal placement solutions. On the other hand, Reinforcement Learning has gained momentum as a tool for decision-making, allowing agents to construct and improve policies using feedback from the environment. In this paper, we leverage Deep Reinforcement Learning (DRL) to perform SFC placement on multiple domains. We devise a hierarchical architecture where the local domain agents and the multi-domain agent are trained using different DRL models to perform SFC and sub-SFC placement while satisfying the SLA requirements.

Service Function Chaining (SFC) refers to the process of steering packets between a set of functions to deliver an end-to-end service. It is considered as one of the enabling technologies for 5G along with Software Defined Networks (SDN) and Network Function Virtualization (NFV). One of the challenges for SFC deployment is the end-to-end orchestration, particularly in a multi-domain scenario where additional issues, such as the lack of visibility and control, and interoperability, need to be taken into account. In this paper, we propose a novel framework that leverages on and extends existing standards in order to perform an end-to-end cross-domain orchestration of SFCs, and ensure the desired forwarding of packets between the domains. A Proof of Concept implementation of our approach is also deployed and evaluated.

The future 5G systems ought to meet diverse requirements of new industry verticals, such as Massive Internet of Things (IoT), broadband access in dense networks and ultra-reliable communications. Network slicing is an important concept that is expected to support these 5G verticals and cope with the conflicting requirements of their respective services. Network slicing allows the deployment of multiple virtual networks, or slices, over the same physical infrastructure as well as supporting on-demand resource allocation to those slices. In this paper, we propose an architecture that will explore how both Network Function Virtualization (NFV) and Software Defined Networking (SDN) may be leveraged to secure a network slice on-demand, addressing the new security concerns imposed to the network management by the flexibility and elasticity support. Our proposed framework aims to ensure an optimal resource allocation that manages the slice security strategy in an efficient way. Moreover, experimental performance evaluations are presented to evaluate the security overhead in virtualized environments.