Exogenous Update Scheduling in the Industrial Internet of Things for Minimal Age of Information

Data freshness is extremely important for real-time applications and generally measured with age of information (AoI). Related studies typically assume that fresh data can be generated at any time. However, in industrial Internet of Things (IIoT) applications, such as alerting, monitoring, or task-oriented operations, data generation is often exogenous and occurs within a finite window. This motivates our analysis, where we investigate AoI-minimizing scheduling for status updates from an IIoT source that generates fresh data only at random intervals. We differentiate between infinite and finite horizons, with the latter being more aligned with IIoT tasks. For each scenario, we examine both agnostic (predefined and unchangeable) and source-aware scheduling, based on the probability of fresh data generation and the duty cycle. We provide a tight bound for source-aware scheduling in the infinite-horizon case and exact expressions for the other scenarios. We assess the increase in AoI from sporadic data generation, finding worst-case factors of 3 for agnostic scheduling and 2 for source-aware scheduling. However, these estimates are pessimistic when the data generation probability is at least an order of magnitude higher than the duty cycle. In such cases, the AoI increase is less than 20% for agnostic scheduling and almost negligible for source-aware scheduling.