Causal Relationship Analysis Using Graph Representations for 5G (CRAGR-5G)

Telecommunications infrastructure has always been at risk of attack, but the proliferation of virtualized network functions may reduce redundancy and resiliency of our telecommunications infrastructure. If virtual function deployments are concentrated within a single cloud provider or zone, an attack on that specific zone could be catastrophic where all or most base stations from all network operators in an area may fail simultaneously.In Phase 1 we will create a model of 5G networks that have virtual network functions disaggregated into data centers. We will identify parameters to model and establish a baseline model to begin identifying and studying critical parameters. Parameters of interest may include a correlation factor between outages so that we can determine the sensitivity of the model to this correlation. This correlation of outages may be due to an outage of a cloud provider who is providing services to many different network operators, or a cyber attack that impacts specific deployments across a subset of mobile network operators, or the correlation may be contained within a single network operator. Another parameter of interest may be the intensity of the outages. We expect outages with high intensity that impact many network operators to be the worst-case scenario. Depending on the coverage area, for lesser intensity outages with lower correlations, we will study which is the dominating factor.