End-to-End Delay Performance Analysis with Variable SINR Estimation Precision Using Stochastic Network Calculus

In 5G-enabled Industrial Internet of Things (IIoT) systems, end-to-end (E2E) deterministic delay performance is paramount for supporting ultra-reliable low-latency communications (uRLLCs). While intelligent algorithms enable more accurate acquisition of wireless channel parameters and real-time state information, existing research predominantly assumes predefined statistical distributions for signal-to-interference-plus-noise ratio (SINR), leaving the critical relationship between instantaneous SINR estimation precision and E2E delay bound tightness insufficiently explored. This paper bridges this research gap through comprehensive analysis of statistical E2E delay performance in uplink transmission scenarios, systematically investigating delay characteristics under varying SINR estimation precision levels. We propose four SINR estimation precision levels and derive closed-form expressions for the effective capacity (EC) of wireless channels in interference-aware communication networks. Furthermore, an upper bound on the E2E delay is analyzed using stochastic network calculus (SNC). Our results demonstrate that the availability of instantaneous signal power information significantly improves the tightness of the E2E delay upper bound by leveraging SNC.