Ensuring QoS in GF-NOMA Networks: Light-Weight Resource Allocation Strategy for IoTs

Beyond 5G (B5G) networks aim to provide massive connectivity and support for low-powered Internet of Things (IoT) networks. In this context, Grant-free NOMA (GF-NOMA) enhances spectral efficiency and minimizes communication overhead with light-weight resource allocation over random channels. However, the GF-NOMA suffers from uplink collisions due to random and uncoordinated process, where all devices (including critical services) face the same collision probability. The uniform treatment increases the likelihood of service degradation for delay-intolerant applications. This paper introduces a novel solution that creates QoS-aware resource pools according to QoS requirements, reducing contention for high-priority devices without adding communication overhead. The proposed approach reduces 10% average collision probability, 10 to 40 seconds average delay for successful transmissions. It achieves up to 10 Mbps gain in average throughput with energy-efficient communication, compared to legacy and state-of-the-art schemes. The proposed QoS-aware resource pooling approach is also adaptable to 6G network slicing applications and open radio access networks, providing an inexpensive framework for handling diverse operators and services.