Samsung Research explores the integration of AI and machine learning (ML) in the physical layer of 6G networks, focusing on semantic audio communication and neural joint source-channel coding and modulation (JSCM). Traditional wireless systems prioritize bit-level accuracy, but semantic communication shifts the focus to preserving the meaning of transmitted information, especially for perceptual signals like audio and speech. This approach is particularly beneficial in resource-constrained environments, such as satellite links, wearable devices, weak-signal cellular conditions, and emergency networks (N. Jiang, B. Yu, F. Sun, and H. Ji, “6G AI/ML for Physical-layer: Part II – Audio Traffic with JSCM,” Samsung Research Blog, May 6, 2026. [Online]. Available: https://research.samsung.com/blog/6G-AI-ML-for-Physical-layer-Part-II-Audio-Traffic-with-JSCM) .
The JSCM framework enables end-to-end optimization of source coding, channel coding, and modulation, allowing the system to learn how to compress speech and make it resilient to transmission impairments. This contrasts with conventional audio communication systems, which follow a layered pipeline of compression, channel coding, and modulation. By jointly learning these functions, JSCM improves efficiency and robustness in delivering perceptual information.
Samsung tested the JSCM concept using a 5G-based link-level simulator and a hardware proof-of-concept (PoC) with Universal Software Radio Peripherals (USRPs). Both evaluations showed that semantic audio communication can maintain similar perceptual quality at roughly 10 dB lower signal-to-noise ratio (SNR) or transmit power than conventional systems [1]. This performance gain is significant for real-world applications where energy efficiency and coverage are critical.
The broader implication of semantic audio communication is that future wireless systems can optimize for user perception rather than just bit fidelity. This shift is particularly valuable in scenarios with limited resources, where it can improve coverage, reduce power consumption, and maintain critical communication during emergencies. As 6G research progresses, AI and ML are becoming core enablers of intelligent, adaptive, and efficient communication systems.
