Traffic-Driven Fast RAW Grouping in Wi-Fi HaLow Heterogeneous Network

In this paper, we consider a large-scale Wi-Fi HaLow heterogeneous network in a real-world Internet of Things (IoT) environment, where numerous devices are distributed around an access point (AP). These devices collect and transmit data using the restricted access window (RAW) mechanism. They exhibit varying traffic characteristics. Moreover, dependencies among the devices often exist. We aim to maximize the overall throughput by adjusting the RAW grouping decision. The heterogeneity of real network data and dependencies between devices make the RAW grouping process more complicated. To overcome this challenge, we propose a novel traffic-driven RAW grouping approach. Specifically, we build a simulation environment based on real IoT data and NS-3. Then, we analyze the traffic characteristics of each IoT device. This analysis allows us to fully explore the dependencies and cooperation relationships among these devices. Hence, we can aggregate devices with these relationships into clusters. Each cluster is then treated as a supernode which is used as a basic unit for RAW grouping. Then, we use proximal policy optimization (PPO) algorithm to optimize the RAW grouping process via interacting with the environment. Numerical results indicate that the proposed traffic-driven algorithm significantly achieves faster convergence and improves grouping efficiency in large-scale heterogeneous networks compared to baselines.