6G-INTENSE: Intent-Driven Native Artificial Intelligence Architecture Supporting Network-Compute Abstraction and Sensing at the Deep Edge

The 6G ecosystem is anticipated to encompass a diverse array of business actors, including mobile network operators, infrastructure providers, and service providers. Existing network management systems (NMSs), which traditionally map technological and administrative domains on a one-to-one basis, are inadequate for managing such a complex environment. This necessitates innovative distributed architectures with varying levels of abstraction. To address these challenges, we present 6G-INTENSE, a novel network management architecture designed to handle the distributed and heterogeneous nature of 6G through two key abstractions. First, the network-compute fabric (NCF) component facilitates multidomain mapping at the resource level by abstracting resource pools, including deep-edge nodes equipped with sensing capabilities. Second, the domain manager and orchestrator (DMO) serves as a central entity for service orchestration, providing a global perspective on the infrastructure. The proposed architecture leverages the intent-based networking (IBN) paradigm at all architectural layers, emphasizing its critical role in enabling autonomous networks. Furthermore, it integrates an innovative native artificial intelligence (AI) toolkit, which is pervasive across all layers.