User-Intent-Driven Semantic Communication via Adaptive Deep Understanding

2025-08-07

AI Summary

Existing semantic communication systems struggle to deeply comprehend and generalize users’ abstract intentions. To address this, we propose an intention-driven semantic communication framework: (1) a user-intention prior generation mechanism built upon multimodal large language models (LLMs) to model high-level semantic intentions; (2) a joint masked-guided attention and channel-state-aware module enabling adaptive semantic feature encoding and robust transmission; and (3) an end-to-end trainable semantic encoder-decoder architecture. Under Rayleigh fading channels at 5 dB SNR, our framework achieves 8%, 6%, and 19% improvements in PSNR, SSIM, and LPIPS, respectively, significantly outperforming DeepJSCC. This work pioneers the integration of LLM-driven intention understanding into semantic communication, marking a paradigm shift from fidelity-oriented transmission to intention-aware meaning delivery.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Multiagent Systems: Agent Communication
• Natural Language Processing: Sentence-level Semantics, Textual Inference, etc.

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Search and Retrieval-Augmented AI: Large language models for search

Abstract

Semantic communication focuses on transmitting task-relevant semantic information, aiming for intent-oriented communication. While existing systems improve efficiency by extracting key semantics, they still fail to deeply understand and generalize users' real intentions. To overcome this, we propose a user-intention-driven semantic communication system that interprets diverse abstract intents. First, we integrate a multi-modal large model as semantic knowledge base to generate user-intention prior. Next, a mask-guided attention module is proposed to effectively highlight critical semantic regions. Further, a channel state awareness module ensures adaptive, robust transmission across varying channel conditions. Extensive experiments demonstrate that our system achieves deep intent understanding and outperforms DeepJSCC, e.g., under a Rayleigh channel at an SNR of 5 dB, it achieves improvements of 8%, 6%, and 19% in PSNR, SSIM, and LPIPS, respectively.

Problem

• Deeply understand and generalize users' real intentions in semantic communication
• Effectively highlight critical semantic regions for intent-oriented communication
• Ensure adaptive, robust transmission across varying channel conditions

Innovation

• Multi-modal large model for intent prior generation
• Mask-guided attention for semantic highlighting
• Channel awareness for adaptive robust transmission