Home/Glossary/6G Glossary/Semantic Communications
6G • RAN

Semantic Communications

Semantic and Goal-Oriented AI Communications (SemCom)

A post-Shannon communication paradigm that transmits the underlying meaning and task-relevant information rather than raw bits.

Technical Explanation

Claude Shannon's classical 1948 information theory focused exclusively on reproducing exact bit sequences over noisy channels. Semantic Communications addresses Weaver's Level B (semantic) and Level C (effectiveness) communication. By training end-to-end Deep Neural Networks (Joint Source-Channel Coding - JSCC) across transmitter and receiver, 6G systems encode the essential semantics required to accomplish a specific task (e.g., classifying an obstacle or executing a remote robotic movement), yielding robust performance at signal-to-noise ratios far below classical Shannon limits.

Key Functions

  • Joint Source-Channel Coding (JSCC) powered by deep neural networks
  • Extraction and transmission of semantic features tailored to specific receiver tasks
  • Extreme bandwidth reduction (>90%) for computer vision, robotics, and speech
  • Unprecedented error resilience operating reliably in extreme low-SNR environments
  • Goal-oriented optimization prioritizing data packets that impact real-world outcomes
Specifications
ITU-R M.2160 (AI & Comms), IEEE JSAC Special Issue on SemCom, 3GPP Rel-20
Interfaces
Sem-UuJSCC-EncodeSemantic-Broker

Related 6G Concepts

Want to memorize 6G concepts like this one?
Study it with SuperMemo SM-2 spaced repetition flashcards.
Practice 6G Flashcards