Semantic Hierarchy-Aware Semantic Communication System for Multi-Stage Implicit Semantic Reasoning

Reasoning-based semantic communication has shown great potential for enhancing transmission efficiency and improve user experience in intelligent services, yet existing methods face challenges in achieving high inference accuracy within large-scale, complex knowledge bases. This letter proposes a multi-stage reasoning-based semantic communication framework that leverages the inherent hierarchical structure of knowledge bases for effective and robust implicit semantic reasoning. By guiding the exploration process via a semantic hierarchy, our approach disentangles complex reasoning tasks into simpler subtasks, improving the accuracy of implicit semantic inference. Experimental results validate our method not only surpasses current state-of-the-art techniques in terms of inference precision but also demonstrates greater robustness under noisy channels.