ISAC-Assisted Local Environment Reconstruction for RAN Digital Twin Synchronization

Abstract

Radio Access Network Digital Twin (RAN-DT) is a digital representation of the radio access network and its related physical environment. A key component of RAN-DT is the propagation-relevant environment representation, which captures structures affecting channel propagation characteristics. To reflect time-varying local radio environments, this paper introduces an integrated sensing and communication (ISAC)assisted local environment reconstruction framework for propagation-relevant RAN-DT synchronization. In the proposed framework, uplink sensing observations are collected through base station (BS)-user equipment (UE) cooperative aperture sensing, where the BS-side receiveport dimension and the UE trajectory dimension jointly form an equivalent two-dimensional synthetic aperture. Based on the collected wideband observations, networkside UE-specific imaging is performed using a weighted and normalized back-projection method, followed by non-coherent fusion of UE-specific reconstructed images to combine multiple sensing views without requiring strict inter-UE phase alignment. A measurementtriggered local update mechanism is further considered to restrict reconstruction and RAN-DT environment updating to affected local regions. Ray-tracing-based simulations with diffuse scattering show that the proposed method can reconstruct propagation-relevant local wall structures, achieving a wall area coverage ratio of $R_{\text {cov }}=84\%$ after multi-UE fusion, corresponding to a relative improvement of 55.6% over the best single-UE result. By providing more accurate local environment reconstruction and revealing local structures that cause observable power delay profile (PDP) variations, the proposed framework can provide useful update information for maintaining accurate propagation characteristics in RAN-DT and improving its real-time effectiveness.