Analysis of Grant-Free Access for URLLC Service: A Stochastic Network Calculus Perspective

The strict low delay and high reliability requirements of Ultra-Reliable and Low-Latency Communication (URLLC) service pose great challenges to the radio access design of mobile communication systems. To meet such requirements, Grant-Free (GF) access is proposed as a promising technology, which effectively reduces the time of requesting/waiting for a grant. However, the dynamic wireless channel environment, random activities of ubiquitous IoT devices, and changeable distributions of packet arrivals make it hard to characterize and analyze the delay and reliability of GF access for URLLC service. In this paper, we develop a novel analytical framework to shed light on the puzzling relationship between delay and reliability of GF transmission for URLLC service from the perspective of Stochastic Network Calculus (SNC). Based on this framework, the statistical performance is characterized by a delay bound and a delay violation probability, and is achieved by exploiting SNC in the Signal-to-Noise Ratio (SNR) domain with the Mellin Transform of packet arrival and service processes. Simulation results validate the theoretical analysis of statistical delay performance with various network parameters.