Interference Aware Ultra Reliable Communications in In-X IoT Sub-Networks

In an environment with strict constraints on maximum permissible delay and a high density of interconnected devices, achieving global system coordination becomes impractical. This lack of coordination results in increased noise due to interference. Under such conditions, interference levels can vary significantly from one packet to another, with only their statistical properties being estimable before transmission. Notably, the variance of the resulting noise remains unpredictable. To address this challenge, we propose an approach to optimize the modulation and coding schemes. Our method models interference using a mixture of exponential distributions. To accurately estimate its parameters, we employ a bootstrap method. We also propose a Quickest Change Detection approach to identify changes in the interference distribution. This approach allows us to determine transmission parameters that ensure a predefined success probability for transmitted packets, without requiring additional listening and estimation tasks on the end devices.