UNITY-6G project partners publish a study on semantic communications in IEEE Communications Standards Magazine

UNITY-6G partners coauthored a new paper titled “Semantic Non-Terrestrial Communications with Open RAN-enabled 6G.” The work by CTTC’s Engin Zeydan, Luis Blanco, Cristian J. Vaca-Rubio, and Marius Caus, and Kapal Dev of the Munster Technological University, presents a novel framework, SEM-NTN, that combines semantic communications with Open RAN and Non-Terrestrial Networks (NTN) to address the performance demands of future 6G systems. By leveraging AI-driven feature extraction, adaptive compression, and intelligent resource allocation, the framework significantly improves bandwidth efficiency and reduces transmission delays. The study demonstrates that SEM-NTN can cut delays by up to 87.5% while maintaining nearly the same AI inference accuracy as traditional approaches, highlighting its potential for scalable, low-latency, and resource-efficient 6G satellite-terrestrial integration.

Abstract — The integration of semantic communication with Open Radio Access Network (O-RAN)-enabled Non-Terrestrial Networks (NTN) is a key enabler for 6G, optimizing bandwidth efficiency, Artificial Intelligence (AI)-native inference and intelligent satellite-terrestrial integration. Unlike traditional bit-level transmission, semantic-aware networks extract and transmit only meaningful information, reducing overhead and improving adaptability. This paper presents SEM-NTN, a semantic-aware O-RAN-enabled NTN framework, that leverages AI-driven feature extraction, adaptive compression, and dynamic resource allocation. Simulation results show that SEM-NTN reduces transmission delay by up to 87.5% while maintaining AI inference accuracy close to that of full-quality compression. Notably, semantic-aware compression achieves up to 84.6% mean Average Precision (mAP) and 77.6% mean Intersection over Union (mIoU) under a 5ms delay constraint—closely matching the uniform baseline (85% mAP, 78% mIoU) but with significantly improved latency performance. These findings highlight SEM-NTN’s potential for scalable, low-latency, and resource-efficient 6G communications. The paper concludes by outlining key challenges in semantic protocol design, real-time adaptation, and standardization for AI-native network integration.

Keywords — semantic communications, non-terrestrial networks, O-RAN, AI-native networks

Explore the full paper to understand the potential of semantic communications for scalable, resource-efficient 6G systems.

more insights