A Non-Uniform Probabilistic Backoff Algorithm in IEEE 802.15.6-Based Wireless Body Area Networks

Abstract

Random access is widely used in IEEE 802.15.6 Wireless Body Area Networks (WBANs) to support heterogeneous and time-sensitive physiological data. While the IEEE 802.15.6 standard and its recent extensions reduce latency and enable quality-of-service (QoS) differentiation by assigning non-overlapping contention windows, they allow nodes to re-enter contention immediately after collisions without delay penalties. This increases the risk of repeated collisions and leads to consecutive packet drops, which can be critical in real-time monitoring systems where continuous data update is essential. To address this limitation, we propose the Generous Non-Uniform Non-Overlapping Backoff Algorithm (GNU-NOBA), which integrates (m, k)-firm constraints and adaptively reshapes the backoff distribution based on recent transmission outcomes. This preserves priority-aware access while guiding failure-prone nodes toward less-contended slots, thereby reducing the probability of successive failures. Simulation results show that GNU-NOBA significantly improves collision avoidance, delivery continuity, and fairness under congested WBAN conditions. © 2025 Elsevier B.V., All rights reserved.