The field of wireless communications is shifting towards semantic-aware transmission, emphasizing task-relevant information over traditional bit-centric approaches. Recent developments focus on adaptive semantic transmission frameworks, integrating techniques like deep reinforcement learning, generative artificial intelligence, and low-rank adaptation to enhance performance in dynamic environments. Noteworthy papers include TOAST, which introduces a unified framework for task-oriented adaptive semantic transmission, and Cross-Attention Message-Passing Transformers, which proposes an AI-native foundation model for unified and code-agnostic decoding in 6G networks. These innovative works demonstrate significant improvements in classification accuracy, reconstruction quality, and coding efficiency, paving the way for more efficient and reliable communication systems.
Semantic Communications and Adaptive Transmission in Dynamic Environments
Sources
TOAST: Task-Oriented Adaptive Semantic Transmission over Dynamic Wireless Environments
On Drug Delivery System Parameter Optimisation via Semantic Information Theory
Reliable Transmission of LTP Using Reinforcement Learning-Based Adaptive FEC
On the Resilience of Underwater Semantic Wireless Communications
Cross-Attention Message-Passing Transformers for Code-Agnostic Decoding in 6G Networks
Multi-User Generative Semantic Communication with Intent-Aware Semantic-Splitting Multiple Access
AI-Empowered Channel Generation for IoV Semantic Communications in Dynamic Conditions
Measurements and Modeling of Air-Ground Integrated Channel in Forest Environment Based on OFDM Signals