Fair and Dynamic Channel Grouping Scheme for IEEE 802.11ah Networks

Abstract

IEEE 802.11ah, marketed as Wi-Fi HaLow, is the new Wireless LAN (WLAN) standard for the Internet of Things (IoT) with many enhancements to the 802.11 standard. One of the new features, mentioned as the Restricted AccessWindow (RAW), focuses on enhancing scalability in extremely dense deployments. RAW divides stations into groups and reduces contention and collisions by permitting channel access to one group at a time. However, the standard does not mandate any optimum RAW grouping strategy. Existing station grouping schemes for enhanced throughput and Quality of Service (QoS) considered a fixed number of groups, though it is one of the important factors in determining the overall throughput. In this paper, we have proposed a real-time grouping scheme non-constrained on the number of groups for homogeneous periodic traffic with the same QoS. The proposed scheme uses Agglomerative Hierarchical Clustering for station grouping where fairness and throughput are used as metrics for distance measure and level selection. Level throughput is estimated from its clusters(groups), using a Neural Network regressor with a 3-dimensional input vector. Evaluation of the network is done after every beacon interval and optimum station grouping and group parameters are broadcasted before the start of a next beacon interval. The proposed model was tested against uniform grouping and random grouping, an improvement of 20%, and 50% in the normalized throughput was observed. Fairness ratio of 0.945 was achieved with the proposed model.