arXiv · 2611.15520
Over-the-Air Computation for Imperfectly Synchronized Multi-View Semantic Communication
Abstract
Compared to classical task-agnostic communication systems, which focus on reliable transmission of a message, semantic communication (SemCom) focuses on transmitting the task-relevant meaning behind the message. In this paper, we design a SemCom system for a distributed multi-view scenario, where multiple semantic devices observe a scene and collaborate through a single semantic receiver to execute a task. We propose adopting an Over-the-Air Computation (AirComp) channel for efficient use of resources, as AirComp uses a superposition of signals to directly compute a function, which in SemCom is decoding the semantic variable. However, AirComp relies heavily on perfect time and phase synchronization, which is not practical in real-world communication. We show that timing offsets of standards like 5G cause negligible degradation, and that larger timing offsets can be compensated by training. Nevertheless, the system cannot deal with the encountered levels of phase synchronization errors, even when trained with those levels of synchronization errors. Therefore, we propose integrating existing phase synchronization algorithms into the AirComp SemCom system training. Simulation results show that the proposed system can approach the task execution performance of a perfectly synchronized SemCom system.
Explore related subjects
Keep this discovery
Explore connections, maps & timelines
Maximilian H. V. Tillmann, Julian Sündermann, Ahmad Halimi Razlighi, Edgar Beck, Armin Dekorsy. 2026-11-11. Over-the-Air Computation for Imperfectly Synchronized Multi-View Semantic Communication. https://arxiv.science/abs/2611.15520
Citations retain the original authors and DOI and include this site’s record URL. If this discovery service contributed to your research, you can also cite arXiv Science by Indiciae: BibTeX · RIS.