This paper introduces a novel framework that integrates agentic ai with ibn to enable autonomous management, configuration, and optimization of mobile network services and resources. Leveraging the advanced reasoning and natural language processing capabilities of an llm, the proposed architecture translates high-level user intents into precise network actions, facilitating user-friendly and scalable network orchestration. The framework employs a distributed multi-agent system, where specialized agents collaborate to decompose user intents, provide computational infrastructure, and deploy services using industry-standard iac tools. By supporting natural language interactions, the system reduces operational complexity and enhances accessibility for users with varying technical expertise. Experimental evaluations demonstrate significant improvements in task completion rates, response accuracy, and operational efficiency compared to traditional manual methods, particularly for complex network management tasks. In essence, this work creates an intelligent network orchestration framework that adapts to user needs by automatically configuring network and computing resources while operating with minimal human intervention.
As the complexity of electric systems increases, so does the required effort for the monitoring and management of grid operations. To solve grid performance issues, smart grids require the exchange of higher volumes of data, high availability of the telecommunication infrastructure, and very low latency. The fifth generation (5G) mobile network seems to be the most promising technology to support such requirements, allowing utilities to have dedicated virtual slices of network resources to maximize the service availability in case of network congestions. Regarding this evolving scenario, this work presents the Smart5Grid project vision on how 5G can support the energy vertical industry for the fast deployment of innovative digital services. Specifically, this work introduces the concept of network applications (NetApps), a new paradigm of virtualization that are envisioned to facilitate the creation of a new market for information technology (IT), small and medium enterprises (SMEs), and startups. This concept, and the open architecture that facilitates its implementation, is showcased by four real-life 5G-enabled demonstrators: (1) automatic fault detection in a medium voltage (MV) grid in Italy, (2) real-time safety monitoring for operators in high voltage (HV) substations in Spain, (3) remote distributed energy resources (DER) monitoring in Bulgaria, and (4) wide area monitoring in a cross-border scenario between Greece and Bulgaria.
As mobile networks evolve beyond 5G, the convergence of cloud, edge and network domains renders siloed orchestration unsustainable. Achieving the adaptability and autonomy required by 6G requires an intelligence layer capable of reasoning, coordination, and continuous learning across distributed systems. While Generative AI (GenAI) and Large Language Models (LLMs) have introduced cognitive automation, the next step is Agentic AI, where autonomous agents reason, negotiate and act collaboratively. This article introduces two key contributions: 1) a blueprint for Domain Agentic Systems (DASs), i.e., composite, scope-bounded subsystems that combine multiple agents, service engines and a communication bus to deliver specific orchestration functionalities, and 2) the AGILE-6G framework, a next-generation orchestration paradigm that embeds an agentic intelligence layer above traditional management systems. AGILE-6G organizes DASs across end-to- end (global) and specialized domains (e.g., Radio Access Network), enabling context-aware decision-making, self-healing and adaptive service management through agent collaboration. Two representative use cases illustrate how AGILE-6G achieves automatic reconfiguration of network and application services, highlighting its potential to realize scalable and trustworthy orchestration in 6G systems.
This paper introduces a Model Context Protocol (MCP)-based framework that enables simplified management and orchestration of UERANSIM containerized deployments for 5G network simulation. Leveraging the standardized MCP interface, the proposed architecture provides specialized tools that handle the complete lifecycle of simulated gNodeB (gNB) and User Equipment (UE) containers, by enabling users to express intents to generative AI models. The framework implements dynamic configuration management with automatic IP address detection and runtime YAML updates, eliminating manual intervention while maintaining precise control over network parameters. By abstracting complex operations behind user-friendly AI chats, the system reduces operational complexity and enables rapid deployment of 5G test environments. Experimental evaluations demonstrate the framework's effectiveness through a practical deployment scenario integrating Open5GS core network with UERANSIM components, validating both functional correctness and operational efficiency. Performance analysis reveals significant improvements in deployment time and configuration accuracy compared to traditional manual approaches, while maintaining flexibility for diverse testing scenarios.
The continuous expansion and transformation of next-gen networks presents a substantial challenge for management and orchestration (MANO) frameworks to efficiently handle the increasing computational demand, particularly at the network's edge. To enhance MANO decision-making processes, it is essential to proactively identify both periodic and aperiodic patterns that may influence network's behavior. This paper presents a computational load forecasting approach designed to provide MANO frameworks with predictive insights into infrastructure conditions. Forecasting leverages machine learning models such as LSTMs and Bi-LSTMs, employing the federated learning family of algorithms, FedOpt, to enable distributed training. Experimental results demonstrate that the proposed federated learning strategy achieves performance comparable to centralized learning methods while significantly reducing the amount of data exchanged between edge nodes and the cloud.