Toward AI-Native RAN: A Lightweight Edge Prototype for Real-Time Environmental Sensing and Remote Actuation

With the development of Internet of Things (IoT) technology, IoT-enabled smart system, as a fundamental component of Artificial Intelligence-Native Radio Access Network (AI-Native RAN), have attracted an increasing attention from academia and industry. To achieve real-time environmental sensing and remote actuation for IoT smart systems, in this paper, based on ZigBee wireless sensor network technology, we design and implement a lightweight edge prototype system by leveraging a "one coordinator, multiple end devices" architecture. Specifically, the Sensing End Devices (SEDs) constitute the perception layer, integrating core data from multiple distributed sensors to monitor parameters such as temperature, humidity, flame, gas concentration, light and human infrared. Additionally, the Actuation End Devices (AEDs) form the execution layer for command response and actuator control. Then, the coordinator node processes these collected data, and uploads them to the Alibaba Cloud IoT Platform through a Wireless Fidelity (Wi-Fi) gateway, enabling remote monitoring of environmental states. By employing an abnormal alarm functional block, the system promptly triggers alerts when a hazardous situation is detected (e.g., fire or gas leaks). Experimental results validate the reliability and effectiveness of the prototype.