Abstract
Cognitive autonomous networks imply that networks will independently derive and execute intelligent decisions, and thereby elevating the human operator's role to a higher level of abstraction. At that level, the operator only specifies the desired outcomes, called intents, which must then be supported by corresponding intent-driven capabilities in the network or its management functions. Although, Intent-Driven Management (IDM) has been published in multiple works, there is still no globally agreed end-to-end view of such IDM systems, let alone a globally agreed definition of intents. This paper provides a comprehensive discussion on the core aspects of IDM systems and combines them into an end-to-end system view with the related example solutions. Contrasting against a short review of related scientific and standards literature, the paper introduces a flexible, generic definition of intents and an end-to-end IDM system architecture as well as the related modeling of intents to support their standardization. The paper also introduces implementation examples fitting the architecture and discusses advanced IDM features that need to be provided, including the ability to detect and resolve conflicts between intents.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have