Digital Twin and Active STAR-RIS Integration for Improved URLLC in Cognitive Radio Networks

This paper introduces a novel integration of Digital Twin (DT) and Active STAR-RIS to enhance URLLC in cognitive radio networks (CRNs). We investigate the uplink data transmission from single-antenna IoT URLLC nodes (UNs) to a multi-antenna base station (BS) facilitated by an active STAR-RIS, focusing on reducing end-to-end latency. An efficient alternating optimization (AO) algorithm is proposed to address the non-convex problem, breaking it into five sub-problems: transmit beamforming, caching and offloading policies, joint communication and computation optimization, and active STARRIS beamforming enhancement. Results show that increasing RIS elements significantly reduces latency through improved signal management and reduced interference. Active STAR-RIS with optimal beamforming outperforms passive and random strategies by dynamically adapting to channel conditions. Higher transmit power and energy levels further decrease latency, with larger data sizes prioritized in resource allocation. The findings highlight the critical role of active STAR-RIS and optimized energy and power levels in minimizing latency and enhancing system performance in CRNs.