Next-generation intent-based networking in B5G with LLMs and explainable AI

Official URL: https://dx.doi.org/10.1007/978-3-032-06418-9_12

As cellular networks evolve beyond 5G (B5G), intent-based networking (IBN) has emerged as a foundational paradigm for achieving autonomous, zero-touch service management. This chapter explores how large language models (LLMs) and explainable AI (xAI) can transform IBN by enabling intuitive intent interpretation, dynamic conflict resolution, and transparent orchestration. We examine the limitations of current rule-based and template-driven approaches and present LLMs as natural language-driven interfaces that bridge the semantic gap between high-level operator goals and low-level network policies. The chapter highlights the unique ability of LLMs to negotiate conflicting intents across multi-agent, multi-domain B5G environments, and to generate human-readable justifications for automated decisions. We also discuss architectural considerations, including integration with platforms like ONAP and O-RAN, and propose a modular system combining LLMs, conflict resolution engines, and xAI techniques such as SHAP, LIME, and counterfactual reasoning. By synthesizing natural language understanding with AI-native orchestration, this chapter charts a path toward scalable, trustworthy, and explainable automation in future communication networks.