Monitoring Under-Modeled Rare Events for URLLC

Immersive and interactive applications that demand ultra-reliable low-latency communication (URLLC) are gaining traction, particularly in the latest 3GPP standards on 5G. The main focus in standards and research has primarily been on protocols based on techniques to combat deep fades, waveform design for short packets, coexistence, etc. However, for ultra-reliability it is imperative to understand, model and combat errors caused by a broader set of events that are not deep fade. In this paper, we consider under-defined rare events, whose behavior cannot be exactly modeled, but can potentially be detected and their frequency monitored, while also building sufficient robustness against them through repetitions in time/frequency and higher SNR. We also describe how persistent attacks such as jamming can be detected in a distributed fashion and how the network can respond to these events in a synchronous way.