Extend structure-aware AoI optimization to broader system settings

Extend the structure-aware online Age of Information optimization framework to time-varying system parameters, interacting queues, and approximate gradient representations.

Background

The paper develops and proves convergence for a structure-aware stochastic-gradient framework that optimizes the update-generation rate when system statistics are unknown. The analysis is established for a single controllable queueing system whose AoI gradient has an exact finite-dimensional representation and whose relevant parameters can be estimated from observations.

The authors explicitly identify extending the framework to time-varying parameters, interacting queues, and settings in which the gradient representation is only approximate as unresolved directions. These extensions would broaden applicability beyond the stationary, single-system, exact-gradient setting analyzed in the paper.