Detecting Targeted Interference in NB-IoT

Many Internet of Things (IoT) applications are considered critical systems, and it is important to guarantee that such deployments are resilient to attacks. An attacker may use radio interference selectively to disrupt communication while minimising the risk of their detection. It is essential to identify such attacks in order to remove the threat. In this work, we propose a novel method of detecting targeted interference in a Narrowband-Internet of Things (NB-IoT) network at the User Equipment (UE). NB-IoT is a recent Low Power Wide Area Network (LPWAN) radio technology used to deploy IoT infrastructures at scale. Network performance data collected at the UE is used to reason about the current interference situation. Subframe loss rates within the downlink channel are monitored and used as input for a statistical anomaly detector. Our evaluation shows that the detector is able to distinguish targeted interference attacks from the impact of naturally occurring interference in cellular networks. This is important as naturally occurring interference requires a different response than targeted interference attacks.