Optimizing Predictive Analytics in 5G Networks Through Zero-Trust Operator-Customer Cooperation

Data availability in softwarized networks plays a fundamental role in various operations, including network function control, management, and orchestration. Despite early trends of designing domain-specific architectures in isolation, interactions between network operators and their customers have often resulted in limited data exchange, and only recently, standardization bodies have addressed this challenge. In this paper, we advocate for a more robust collaboration between operators and customers by introducing a zero-trust analytics service. This service enables the creation of tailored models for the different customers of network operators. We outline the necessary procedures to support such analytics and present a use case that demonstrates how specific operator-provided analytics (network flow detection) can be enhanced through the incorporation of external signals from a customer.