Topology-aware cognitive self-protection framework for automated detection and mitigation of security and privacy incidents in 5G-IoT networks

Internet of Things (IoT) coupled with 5G networks enable unprecedented levels of scalability and performance in the computing industry. These enhanced performance features allow to offer and deploy a wide range of new use cases and services in scenarios such as Smart Cities, Smart Grid or Industry 5.0 just to mention a few. However, the inherent complexity of such networks is a serious concern in terms of security. Furthermore, the vulnerability and low-power constraints of IoT devices make such networks a targeted vector for cyber criminals. In this contribution, authors present an innovative topology-aware Cognitive Self-protection framework able to detect and mitigate attacks in an autonomous way with no human intervention in the wired segments of 5G-IoT multi-tenant networks. Preliminary tests carried out on a realistic emulated testbed show promising results in terms of time spent in stopping DDoS attacks (less than 47 seconds) and scalability for scenarios with different number of tenants and UEs (2 virtual tenants deployed in 4 Edge nodes and up to 64 IoT devices or sensors connected to the infrastructure).