Named Data Networking (NDN) for Collective Network Intelligence: a Smart Museum case study

Future smart systems are increasingly characterized by pervasive AI deployed both at the back- and front-end. However, limiting intelligence to the software layer is insufficient to meet the scalability, adaptability, and resilience requirements of emerging large-scale cyber-physical ecosystems. Indeed, intelligence must extend across both the entire architectural stack and the infrastructure, including the network layer itself. In this paper, we explore Named Data Networking (NDN) as an enabler of collective intelligence at the network level. Unlike traditional IP-based architectures, NDN natively supports in-network caching, name-based routing, and stateful forwarding, thus enabling distributed decision-making mechanisms to emerge directly from the network fabric. We argue that these properties can be interpreted as forms of collective intelligence, where the network collaboratively optimizes content dissemination, data retrieval, and resource utilization without centralized orchestration. We analyze how NDN primitives inherently support adaptive behavior, context awareness, and distributed knowledge aggregation, by presenting a Smart Museum-related case study as an example. Furthermore, we discuss architectural extensions that enhance NDN's capability to support cross-layer intelligence spanning devices, network, and applications.