Abstract
We investigate the performance of uplink communications in Internet-of-Things (IoT) networks under passive attacks from eavesdroppers capable of intercepting individual links. An industrial IoT network is studied where multiple sensor classes, each with a dedicated local access point (LAP), are connected to one or more small cell access points (SAP) which in turn are connected to a central server. The server interfaces the IoT network to industrial sensor monitoring systems which periodically request sensor readings. We propose a unique attack resilient IoT sensor reporting model and study the performance of this system under strict latency as well as secrecy constraints. We show through simulation results how the proposed model behaves in different scenarios and compare the resilience to attack capability for the different cases.