A DoS Attack Against 5G-Based IIoT Networks Utilizing Standard Network Slicing Mechanisms

5G networks promise to integrate several new technologies including cloud-based virtualization of physical resources and network functions, as well as the separation of the control plane from the data plane. Network slicing-the ability to have multiple virtual networks simultaneously share the same physical infrastructure-plays a central role in these technologies. Network slicing, however, opens the door to new attacks from malicious network tenants. In this work, we show how a side-channel that leaks information about one network tenant to another can be created and used to launch a Denial of Service Attack (DoS). The attack involves three main steps: a reconnaissance phase to obtain the victim network’s traffic patterns over time by exploiting a side channel based on standard 5G network slicing mechanisms; training a Long Short-Term Memory (LSTM) model to be able to predict the victim network’s peak traffic times; and launching the DoS attack during those peak periods, again via standard 5G network slicing mechanisms. A successful attack on a typical Industrial IoT (IIoT) smart manufacturing application was demonstrated using OMNeT++ simulations and real IoT-23 traffic traces.