This paper focuses on a graph-based approach to monitor the core network, including the backbone and
the CRANs, to improve customer experience. The backbone sites consist of three national data centers (NDCs) and three regional data centers (RDCs), and within these sites there are core routers, route reflectors, and dozens of peering edges that are interconnected. The backbone is integral to the IP infrastructure, connecting to 28 CRANs, where each CRAN features a pair of large aggregate routers (AR), Xfinity aggregation router (XAR), residential U routers (RUR), and commercial super U ring routers (SUR), where U refers to the shape of the surrounding topology. Depending on the needs, CRANs can extend several layers deep with multiple RURs and residential edge routers (RERs). The physical layer, also known as the optical/transport layer, connects these sites, while the network layer features logical connections. This complex mesh of devices, services, protocols, and connections, with built-in redundancies, makes event detection and localization particularly challengin