Implementation of an Age Prioritized Wi-Fi Network with Multi-Armed Bandit Polling

This paper presents a method for prioritizing transmissions considering the Age of Information (AoI) metric in communication systems that are compatible with the IEEE 802.11 standard. The proposed method includes Last-In-First-Out (LIFO) at users and a multiuser scheduling algorithm that uses a Multi-Armed Bandit (MAB) model at the access point. The system was implemented on ESP32 microcontrollers and Raspberry Pi devices to evaluate performance, making our study one of the few real-life implementations of age-aware WiFi. The performance of the proposed method is compared with that of the WiFresh method, which uses a Maximum Weight matching scheduler. The simulation and implementation results indicate that the proposed method improves the Peak and mean AoI in the system compared to Wi-Fi. The implementation demonstrates the applicability of the method in real-world scenarios.